GHG NatioNal iNveNtory report South africa · 2018-11-07 · 6 GHG National inventory report for...

2000 - 2012

GHG NatioNal iNveNtory reportSouth afr ica

environmental affairs Environmental Affairs Department:

REPUBLIC OF SOUTH AFRICA

IMPRINT

Published by

Department of environmental affairs (Dea)

General responsibility

Jongikhaya Witi

Individual chapters:

Executive summary - luanne Stevens, Jongikhaya Witi

Chapter 1: Introduction - luanne Stevens, Jongikhaya Witi

Chapter 2: Trends in GHG Emissions - luanne Stevens, Jongikhaya Witi

Chapter 3: Energy Sector - lungile Manzini

Chapter 4: Industrial Processes and Product Use - Jongikhaya Witi

Chapter 5: Agriculture, Forestry and Land Use - luanne Stevens, lindeque Du toit

Chapter 6: Waste - Jongikhaya Witi

Report compiled by

luanne Stevens

Report edited by

phindile MangwanaDumisani Nxumalo

External report reviewers:

IBIS Environmental Social Governance Consulting South Africa Pty Ltd (IBIS) Generalists - paul Ben israel and Simon Clarke

Energy Sector - anthony Dane

Industrial Processes and Product Use (IPPU) Sector - Simon Clarke

Agriculture Forestry and other Land Uses (AFOLU) Sector - tony Knowles

Waste Sector - Fred Goede

Layout by

twaai Design

Contact information

Department of environmental affairsenvironment House473 Steve Biko Streetarcadiapretoria 0001South africa

tel: +27 12 399 9148email: [email protected]

published in South africa - august 2017

this report was kindly funded through the GiZ implemented Climate Support programme, which is part of the international Climate initiative (iKi). the Federal Ministry for the environment, Nature Conservation, Building and Nuclear Safety (BMUB) supports the iKi on the basis of a decision adopted by the German Bundestag.

GHG National Inventory Report for South Africa 2000 -2012 3

GHG NATIONALINVENTORY REPORT

South africa2000 - 2012

January 2017

environmental affairs Environmental Affairs Department:

REPUBLIC OF SOUTH AFRICA

GHG National Inventory Report for South Africa 2000 -20124

PREFACE

this report has been compiled for the Department of environmental affairs (Dea) in response to South africa’s obligation to report its greenhouse (GHG) emissions to international climate change bodies. the report is prepared in accordance with the United Nations Framework Convention on Climate Change (UNFCCC). this inventory was compiled by making use of the intergovernmental panel on Climate Change (ipCC) 2006 excel spread sheet guidelines and the Good practice Guidance.

preface

this report is published by the Dea, South africa. an electronic version will be available on the website of the Department of environmental affairs (http://www.saaqis.org.za/) once the review process is completed. information from this report may be reproduced, provided the source is acknowledged.


ACKNOWLEDGEMENTS

Many people and institutes were involved in the compilation of this National inventory report for 2012. the main information on the energy and industrial processes and product use (ippU) sectors was provided by the Department of energy (Doe), the Department of Mineral resources (DMr), Business Unity South africa (BUSa) members, the Chamber of Mines, eskom, Sasol, and petroSa. the agriculture, forestry and other land use (aFolU) sector was prepared by North West University (NWU) and Gondwana environmental Solutions, with the major data suppliers being tshwane University of technology (tUt), the Department of environmental affairs, the Department of agriculture, Forestry and Fisheries (DaFF), Geoterraimage (Gti), the Food and agriculture organization (Fao), Forestry Sa and the agricultural research Council (arC). the waste sector

was compiled by the Dea with the main data providers being Statistics South africa (StatsSa), the Department of environmental affairs, the World resources institute (Wri) and the Council for Scientific and industrial research (CSir).

We greatly appreciate all the contributions from organizations and individuals who were involved in the process of completing this Nir. Special thanks to the Department for international Development (DFiD) for providing funding for the land change mapping and updating of the agriculture, forestry and land-use sector, as well as funds for the compilation of this national inventory report (Nir). We would also like to thank all reviewers of the various sector sections as well as the reviewers of the completed Nir.


AUTHORS AND CONTRIbUTORS

GeNeral reSpoNSiBility:

Jongikhaya Witi

iNDiviDUal CHapterS:

executive summary: Luanne Stevens, Jongikhaya Witi

Chapter 1: introduction - Luanne Stevens, Jongikhaya Witi

Chapter 2: trends in GHG emissions - Luanne Stevens, Jongikhaya Witi

Chapter 3: energy Sector - Lungile Manzini

Chapter 4: industrial processes and product Use - Jongikhaya Witi

Chapter 5: agriculture, Forestry and land Use - Luanne Stevens, Lindeque Du Toit

Chapter 6: Waste - Jongikhaya Witi

report CoMpilatioN:

Luanne Stevens

report eDitor:

Phindile Mangwana and Dumisani Nxumalo

exterNal report revieWerS:

iBiS environmental Social Governance Consulting South africa pty ltd (iBiS) Generalists – Paul Ben Israel and Simon Clarke; energy Sector – Anthony Dane, industrial processes and product Use (ippU) Sector – Simon Clarke; agriculture Forestry and other land Uses (aFolU) Sector – Tony Knowles; and Waste Sector – Fred Goede.

authors and Contributors


CONTENTS

PREFACE 4

ACKNOWLEDGEMENTS 5

AUTHORS AND CONTRIbUTORS 6

LIST OF AbbREVIATIONS 21

UNITS, FACTORS AND Abb REVIATIONS 25

ExECUTIVE SUMMARY 26

eS1 Background information on South africa’s GHG inventories 26

eS2 Summary of South africa’s GHG emissions 31

eS3 South africa’s indicators 34

eS4 other information 34

eS5 Conclusions and recommendations 41

1 INTRODUCTION 42

1.1 Climate change and GHG inventories 42

1.2 Country background 42

1.2.1 National circumstances 42

1.2.2 institutional arrangements for inventory preparation 45

1.3 inventory preparation 48

1.3.1 Data collection procedures and plans 48

1.3.2 Data archiving and storage 49

1.3.3 Brief description of methodologies and data sources 51

1.3.4 Brief description of key categories 56

1.4 information on Qa/QC plan 66

1.4.1 Quality control 66

1.4.2 Quality assurance (Qa) 67

1.5 evaluating uncertainty 69

1.6 General assessment of completeness 69


2 TRENDS IN GHG EMISSIONS 72

2.1 trends for aggregated GHG emissions 72

2.2 emission trends by gas 74

2.2.1 Carbon dioxide 74

2.2.2 Methane 75

2.2.3 Nitrous oxide 76

2.2.4 Fluorinated gases 77

2.3 Emissiontrendsspecifiedbysourcecategory 78

2.4 emission trends for indirect GHG 79

3 ENERGY SECTOR 80

3.1 an overview of the energy sector 80

3.1.1 energy demand 80

3.1.2 energy reserves and production 80

3.1.3 transport 82

3.2 GHG emissions from the energy sector 83

3.2.1 overview of shares and trends in emissions 83

3.2.2 Key sources 88

3.3 Fuel combustion activities [1a] 88

3.3.1 Comparison of the sectoral approach with the reference approach 89

3.3.2 Feed stocks and non-energy use of fuels 89

3.3.3 energy industries [1a1] 89

3.3.4 Manufacturing industries and construction [1a2] 98

3.3.5 transport [1a3] 101

3.3.6 other sectors [1a4] 111

3.3.7Non-specified[1A5] 116

3.4 Fugitive emissions from fuels [1B] 118

3.4.1 Solid fuels [1B1] 118

3.4.2 oil and natural gas [1B2] 121

3.4.3 other emissions from energy production [1B3] 122

Contents


4 INDUSTRIAL PROCESSES AND OTHER PRODUCT USE 124

4.1 an overview of the ippU sector 124



4.1.3 Completeness 125

4.2 Mineral production [2a] 128

4.2.1 Source category description 128


4.2.3 Methodological issues 131

4.2.4 Data sources 131

4.2.5 Uncertainty and time-series consistency 133

4.2.6SourcespecificQA/QCandverification 133

4.2.7Source-specificrecalculations 133

4.2.8Source-specificplannedimprovementsandrecommendations 134

4.3 Chemical industry [2B] 134


4.3.2 trends in emissions 135




4.3.6Source-specificQA/QCandverification 138



4.4 Metal industry [2C] 138



4.4.3 Methodology 141







4.5 Non-energy use of fuels and solvent use [2D] 145

4.5.1 Source-category description 145








4.6 production uses as substitutes for ozone-depleting substances [2F] 147








5 AGRICULTURE, FORESTRY AND OTHER LAND USE 149

5.1 overview of the sector 149

5.2 GHG emissions from the aFolU sector 150



5.2.3 recalculations 152

5.3 livestock [3a] 152


5.3.2 overview of trends in activity data 152

5.3.3 enteric fermentation [3a1] 154

5.3.4 Manure management [3a2] 165

Contents


5.4 land [3B] 174

5.4.1 overview of the sub-sector 174

5.4.2 overview of shares and trends 175

5.4.3 General methodology 178

5.4.4 Method for obtaining land-cover matrix 180

5.4.5Landusedefinitionsandclassifications 185

5.4.6 recalculations 188

5.4.7 Forest land [3B1] 189

5.4.8 Cropland [3B2] 201

5.4.9 Grassland [3B3] 205

5.4.10 Wetlands [3B4] 207

5.4.11 Settlements [3B5] 209

5.4.12 other land [3B6] 210

5.5 aggregated sources and non-Co2 emission sources on land [3C] 212


5.5.2 Biomass burning [3C1] 212

5.5.3 liming and urea application [3C2 and 3C3] 218

5.5.4 Direct N2o emissions from managed soils [3C4] 220

5.5.5 indirect N2o emissions from managed soils [3C4] 226

5.5.6 indirect N2o emissions from manure management [3C6] 227

5.5.7 Harvested wood products 229

6 WASTE SECTOR 232

6.1 overview of sector 232

6.2 overview of shares and trends in emissions 232

6.3 Key categories 233

6.4 Solid waste disposal on land [4a] 233

6.4.1 Source-category description 233









6.5 Wastewater treatment and discharge [4D] 236





6.5.5 Uncertainties and time-series consistency 238




7 REFERENCES 241

8 APPENDIx A: SUMMARY TAbLES 252

9 APPENDIx b: KEY CATEGORY ANALYSIS 278

9.1 level assessment: 2000 (including FolU) 278

9.2 level assessment: 2000 (excluding FolU) 282

9.3 level assessment: 2012 (including FolU) 286

9.4 level assessment: 2012 (excluding FolU) 290

9.5 trend assessment: 2000 – 2012 (including FolU) 294

9.6 trend assessment: 2000 – 2012 (excluding FolU) 299

10 APPENDIx C: UNCERTAINTY ANALYSIS 304

10.1 energy sector uncertainty 304

10.2 ippU sector uncertainty 316

Contents


11 APPENDIx D: ENERGY SECTOR – COMPARISON OF THE REFERENCE VS SECTORAL APPROACH 324

12 APPENDIx E: AGRICULTURAL ACTIVITY DATA 332

12.1 livestock sub-categories for sheep, goats and swine 332

12.2 livestock population data 334

13 APPENDIx F: LAND SECTOR ACTIVITY DATA 340

13.1 land cover by climate 340

13.2 land cover by soil type 341

13.3 Biomass and soil data 342

14 APPENDIx G: METHODOLOGY FOR LAND COVER AND LAND USE CHANGE MATRIx 345

14.1 South african 2013-14 National land-cover dataset 345

14.2 1990 – 2013-14 land use change report 385


LIST OF FIGURES

FigureA: InformationflowinNAEIS. 27

Figure B: overview of the phases of the GHG inventory compilation and improvement process. 29

Figure C: institutional arrangements for the compilation of the 2000 – 2012 inventories. 30

Figure D: GHG emissions for South africa between 2000 and 2012 by sector with FolU excluded (top) and included (bottom). 32

Figure e: the highest emitting categories in the South african GHG inventory (excl. FolU) in 2000 (top) and 2012 (bottom). 33

Figure F: Gas contribution to South africa’s GHG emissions between 2000 and 2012. 33

Figure G: South africa’s emissions intensity between 2000 and 2012 (Department of environmental affairs, 2014) 35

Figure 1.1: GHG inventory data collection procedure 49

Figure1.2:ConfigurationandfilescontainedintheGHGinventorydatabase. 50

Figure 1.3: the independent review process for the 2000 – 2012 inventory. 68

Figure 2.1: Greenhouse gases: trend and emission levels (excl. FolU), 2000 – 2012. 73

Figure 2.2: Greenhouse gases: trend and emission levels (including FolU sub-sectors), 2000 – 2012. 73

Figure 2.3: Co2: trend and emission levels of sectors (excl. FolU), 2000 – 2012. 74

Figure 2.4: CH4: trend and emission levels of sectors (excl. FolU), 2000 – 2012 75

Figure 2.5: N2o: trend and emission levels of sectors, 2000 – 2012. 76

Figure 2.6: F-Gases: trend and emission levels of sectors, 2000 – 2012. 77

Figure 2.7: total GHG: trend and levels from sectors (excl. FolU), 2000 – 2012. 78

Figure 3.1: Sector 1 energy: trend in primary energy consumption in South africa, 2000 – 2010. 81

Figure 3.2: Sector 1 energy: average contribution of source categories to total energy sector GHG emissions, 2000 and 2012. 84

Figure 3.3: Sector 1 energy: trend and emission levels of source categories, 2000 – 2012. 85

Figure 3.4: Sector 1 energy: annual change in total GHG emissions between 2000 and 2012. 85

Figure 3.5: Sector 1 energy: trend and emission levels of total GHG’s from the energy sector, 1990 – 2012. 87

Figure 3.6: Sector 1 energy: trend in sources of total GHG emissions from fuel used in the manufacturing industries and construction category (1a2), 2000 - 2012. 98

Figure 3.7: Sector 1 energy: percentage contribution of the various fuel types to fuel consumption in the road transport category (1a3b), 2000 – 2012. 102

list of Figures


Figure 3.8: Sector 1 energy: trend in total GHG emissions from the transport sector, 2000 – 2012. 103

Figure 3.9: Sector 1 energy: percentage contribution of Co2, CH4 and N2o from the transport sector, average over 2000 – 2012 period. 104

Figure 3.10: Sector 1 energy: Fuel consumption in the commercial/institutional category, 2000 – 2012. 112

Figure 3.11: Sector 1 energy: trend in fuel consumption in the residential category, 2000 – 2012. 112

Figure3.12:Sector1Energy:Trendinfuelconsumptionintheagriculture/forestry/fishingcategory, 2000 – 2012. 113

Figure 3.13: Sector 1 energy: trend in total GHG emissions from other sectors, 2000 – 2012. 113

Figure 4.1: Sector 2 ippU: trend and emission levels of source categories, 2000 – 2012. 125

Figure 4.2: Sector 2 ippU: trend and emission levels of the various greenhouse gases, 2000 – 2012. 126

Figure 4.3: Sector 2 ippU: trend and emission levels in the mineral industries, 2000 – 2012. 130

Figure 4.4: Sector 2 ippU: trend and emission levels in the chemical industries between 2000 and 2012. 136

Figure 4.5: Sector 2 ippU: percentage contribution from the various industries and gases to the total accumulated GHG emissions from the metal industries between 2000 and 2012. 140

Figure 4.6: Sector 2 ippU: trend and emission levels in the metal industry, 2000 – 2012. 141

Figure 4.7: Sector 2 ippU: trend and emission levels from the non-energy products from fuels and solvent use category, 2000 – 2012. 146

Figure 4.8: Sector 2 ippU: trends and emission levels of HFC’s, 2005 – 2012. 148

Figure 5.1: Sector 3 aFolU: percentage contribution of the various GHG to the total aFolU inventory, 2000 – 2012. 151

Figure 5.2: Sector 3 aFolU: trends in GHG emissions from the livestock category, 2000 – 2012. 153

Figure 5.3: Sector 3 aFolU - livestock: trends in livestock population numbers, 2000 – 2012. 153

Figure 5.4: Sector 3 aFolU - livestock: trend and emission levels of enteric fermentation emissions in the livestock categories, 2000 – 2012. 155

Figure 5.5: Sector 3 aFolU - livestock: trend in cattle herd composition, 2000 – 2012. 155

Figure 5.6: Sector 3 aFolU – livestock: Contribution of the livestock categories to the enteric fermentation emissions in 2012. 156

Figure 5.7: Sector 3 aFolU – livestock: total manure management trend and emission levels from source categories, 2000 – 2012. 166

Figure 5.8: Sector 3 aFolU – livestock: Manure management CH4 trend and emission levels from source categories, 2000 – 2012. 166


Figure 5.9: Sector 3 aFolU – livestock: Manure management N2o trend an emission levels from source categories, 2000 – 2012. 167

Figure 5.10: Sector 3 aFolU - land: Comparison of land areas and the change between 1990 and 2013-2014. 176

Figure 5.11: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the land sub-sector between 2000 and 2012, differentiated by sub-category. 177

Figure 5.12: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the land sub-sector between 2000 and 2012, differentiated by source category. 177

Figure 5.13: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the forest land between 2000 and 2012, differentiated by sub-category. 190

Figure 5.14: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the forest land between 2000 and 2012, differentiated by source category. 190

Figure 5.15: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in Croplands between 2000 and 2012, differentiated by sub-category. 200

Figure 5.16: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in Croplands between 2000 and 2012, differentiated by source category. 200

Figure 5.17: Sector 3 aFolU – land: trend in CH4 emissions (in Gg Co2eq) in wetlands between 2000 and 2012. 208

Figure 5.18: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission levels, 2000 – 2012. MM = Manure management; MS = Managed soils; BB = Biomass burning. 212

Figure 5.19: Sector 3 aFolU – aggregated and non-Co2 sources: Contribution of the various land categories to the biomass burning emissions, 2000 – 2012. 213

Figure 5.20: Sector 3 aFolU – aggregated and non-Co2 sources: trends and emission levels from liming and urea application, 2000 – 2012. 218

Figure 5.21: Sector 3 aFolU – aggregated and non-Co2 sources: annual amount of lime and urea applied to soils, 2000 – 2012. 219

Figure 5.22: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission levels of direct N2o from managed soils, 2000 – 2012 221

Figure 5.23: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission level estimates of indirect N2o losses from managed soils, 2000 – 2012. 226

Figure 6.1: Sector 4 Waste: trends and emission levels of source categories, 2000 – 2012. 233

list of Figures


LIST OF TAbLES

table a: trends in the national GHG emissions, including and excluding FolU, between 2000 and 2012. 31

TableB: TrendsandlevelsinGHGemissionsfor2000and2012classifiedbysector 31

table C: activities in the 2012 inventory which were not estimated (Ne), included elsewhere (ie) or not occurring (No). 36

table D: recalculated sector GHG emission estimates for 2000 and 2010 for South africa. 38

table e: Dea driven GHGip projects 39

table F: Donor funded GHGip projects 40

table 1.1: the GDp percentage growth in South africa between 2000 and 2012 (Statistics Sa, 2014). 43

table 1.2: institutional arrangements in the energy sector 46

table 1.3: institutional arrangements in the ippU sector 46

table 1.4: institutional arrangements in the aFolU sector 47

table 1.5: institutional arrangements in the waste sector 47

table 1.6: tier method (tM) and emission factor (eF) used in this inventory in the estimation of the emissions from the various sectors. 52

table 1.7: Global warming potential (GWp) of greenhouse gases used in this report (Source: ipCC 2001). 54

table 1.8: Summary of South africa’s key categories for 2012 level assessment and the trend assessment for 2000 to 2012 (including and excluding FolU activities). 57

table 1.9: level assessment results for 2012, excluding FolU contributions. only key categories are shown. 59

table 1.10: level assessment results for 2012, including FolU contributions. only key categories are shown. 60

table 1.11: trend assessment results for 2012 (with 2000 as the base year), excluding FolU contributions. only the key categories are shown. 62

table 1.12: trend assessment results for 2012 (with 2000 as the base year), including FolU contributions. only the key categories are shown. 64

table 1.13: QC activity and procedures applied in this inventory 66

table 1.14: activities in the 2012 inventory which are not estimated (Ne), included elsewhere (ie) or not occurring (No). 70

table 2.1: trends and levels in GHG emissions for South africa between 2000 and 2012. 79

table 2.2: precursor GHG: trend and emission levels in Co and Nox (Gg of gas) from biomass burning, 2000 – 2012. 79

table 3. 1: Sector 1 energy: Contribution of the various sources to total energy GHG emissions. 86


table 3.2: Sector 1 energy: Summary of electricity users in South africa (Source: Doe, 2009a). 90

table 3.3: Sector 1 energy: Net electricity generation capacity and associated consumption (Source: eskom, 2005, 2007, 2011). 91

table 3.4: Sector 1 energy: Summary of GHG emissions from the public electricity producer. 92

table 3.5: Sector 1 energy: Summary of GHG emissions from auto electricity producers, 2000 - 2012. 93

table 3.6: Sector 1 energy: Summary of consumption of fuels and GHG emissions in the petroleum refiningcategory(1A1b),2000-2012. 94

table 3.7: Sector 1 energy: Contribution of Co2, CH4 and N2o to the total emissions from the manufacture of solid fuels and other energy industries category (1a1c), 2000 - 2012. 94

table 3.8: Sector 1 energy: emission factors for GHG emissions (Source: 2006 ipCC Guidelines, vol 2 and Zhou et al., 2009). 95

Table3.9: Sector1Energy:EmissionfactorforthecalculationofGHGemissionsfrompetroleumrefining (Source: 2006 ipCC Guidelines). 96

table 3.10: Sector 1 energy: Fuel consumption (tJ) in the manufacturing industries and construction category, 2000 – 2012. 99

table 3.11: Sector 1 energy: emission factors used in the manufacturing industries and construction category (Source: 2006 ipCC Guidelines). 100

table 3.12: Sector 1 energy: Summary of GHG emissions from international aviation (international bunkers), 2000 – 2012. 106

Table3.13:NetCalorificValuesforthetransportSector(Source:DoE2009a) 106

table 3.14: Sector 1 energy: Fuel consumption (tJ) in the transport sector, 2000 – 2012. 108

table 3.15: Sector 1 energy: Fuel consumption (tJ) in the international aviation category, 2000 – 2012. 108

table 3.16: Sector 1 energy: emission factors used for transport sector emission calculations (Source: 2006 ipCC Guidelines). 109

table 3.17: Sector 1 energy: emission factors used for all other sectors (Source: 2006 ipCC Guidelines). 115

Table3.18:Sector1Energy:TrendinconsumptionandGHGemissionsfromthenon-specifiedsector, 2000 – 2012. 117

Table3.19:Sector1Energy:Emissionfactorsforcalculatingemissionsfromthenon-specifiedsector (Source: 2006 ipCC Guidelines). 118

table 3.20: Sector 1 energy: Fugitive emissions from coal mining for the period 2000 to 2012. 119

table 3.21: Sector 1 energy: Coal mining activity data for the period 2000 to 2012. 120

Table3.22:Sector1Energy:Comparisonofcountry-specificandIPCC2006defaultemissionfactors for coal mining. 121

Table3.23:Sector1Energy:TotalGHGemissionsfromventingandflaringfortheperiod2000–2012. 122

list of tables


table 3.24: Sector 1 energy: total GHG emissions from the category other emissions from energy production (1B3), 2000 – 2012. 123

Table4.1: Sector2IPPU:Classificationofcategoriesofemissionsexcludedfromthisinventory. 127

table 4.2: Sector 2 ippU: activity data for cement, lime and glass production, 2000 – 2012. 132

table 4.3: Sector 2 ippU: Clinker fraction for the period 2000 – 2012. 132

table 4.4: Sector 2 ippU: activity data for the various metal industries, 2000 – 2012. 142

Table4.5: Sector2IPPU:Comparisonofthecountry-specificemissionfactorsforironandsteel production and the ipCC 2006 default values (Source: iron and steel Companies; ipCC 2006 Guidelines). 143

table 4.6: Sector 2 ippU: emission factors for ferroalloy production (Source: 2006 ipCC Guidelines). 144

table 4.7: Sector 2 ippU: total fuel consumption in the non-energy use of fuels and solvent use category, 2000 – 2010. 146

table 5.1: Sector 3 - aFolU: trends in emissions and removals (Gg Co2eq) from the aFolU sector, 2000 – 2012. 150

table 5.2: Sector 3 aFolU – livestock: livestock categories used in the determination of livestock emissions. 159

table 5.3: Sector 3 aFolU – livestock: livestock weights, methane eF’s and ipCC 2006 default eF’s. 161

table 5.4: Sector 3 aFolU – livestock: Weights and methane eF’s for game animals. 163

table 5.5: Sector 3 aFolU – livestock: Comparison between the productivity data used and eF calculated in this inventory and the ipCC 2006 Guideline default values. 164

table 5.6: Sector 3 aFolU – livestock: Manure management system usage (%) for different livestock categories, 2000 – 2012 (Source: DaFF, 2010; Du toit et al., 2013a and 2013b). 171

table 5.7: Sector 3 aFolU – livestock: CH4 manure eF’s for the various livestock compared with the ipCC 2006 default factors and the australian eF’s (Source: Du toit et al., 2013a, b, c; ipCC 2006 Guidelines; aNir, 2009). 172

table 5.8: Sector 3 aFolU – land: the six ipCC land classes and the South african sub-categories within each land class. 179

table 5.9: Sector 3 aFolU – land: land-use matrix for the change between 1990 and 2013-14 (ha) with highlighted numbers on the diagonal showing the area (ha) remaining in the same category and the other cells show the relevant land-use changes (including the transitions since 1990). 181

table 5.10: Sector 3 aFolU – land: annual land cover changes between the various land classes. 182

table 5.11: Sector 3 aFolU – land: Comparison between the land areas from the 2010 inventory and the updated areas in this 2012 inventory. 184


table 5.12: Sector 3 aFolU – land: annual wood harvest data (m3/yr) from plantations between 2000 and 2012 (source: FSa, 2012). 195

table 5.13: Sector 3 aFolU – land: annual removal of fuelwood (m3/yr) from plantations between 2000 and 2012 (source: FSa, 2012). 196

table 5.14: Sector 3 aFolU – land: relative stock change factors (± error) for different management activities on croplands (Source: Department of environmental affairs, 2015; Moeletsi et al., 2013). 198

table 5.15: Sector 3 aFolU – land: above ground biomass for the various land classes. 204

table 5.16: Sector 3 aFolU – aggregated and non-Co2 sources: percentage area burnt within each land class between 2000 and 2012. 215

table 5.17: Sector 3 aFolU – aggregated and non-Co2 sources: Fuel density and combustion fractions for the various vegetation classes. 216

table 5.18: Sector 3 aFolU – aggregated and non-Co2 sources: emission factors applied to the various vegetation classes. 217

table 5.19: Sector 3 aFolU – aggregated and non-Co2 sources: limestone, dolomite and urea use between 2000 and 2012 (source: Fertilizer Society of Sa; FaoStat). 220

table 5.20: Sector 3 aFolU – aggregated and non-Co2 sources: indirect emissions of N2o (Gg Co2eq) due to volatilization from manure management between 2000 and 2012. 228

table 5.21: Sector 3 aFolU – aggregated and non-Co2 sources: Default values used for N loss due to volatilization of NH3 and Nox from manure management (%). the value in the brackets indicates the range. 229

table 5.22: Sector 3 aFolU – other: trend in the HWp Co2 sink (Gg) between 2000 and 2012. 229

table 6.1: Sector 4 Waste: CH4 and N2o emissions from domestic and industrial wastewater treatment, 2000 – 2012. 237

table 6.2: Sector 4 Waste: emission factors for different wastewater treatment and discharge systems (Source: Department of environmental affairs and tourism, 2009). 239

table 6.3: Sector 4 Waste: Distribution and utilization of different treatment and discharge systems (Source: Department of environmental affairs and tourism, 2009). 239

list of tables


aFolU agriculture, Forestry and other land Use

aGB above-ground biomass

Bbl/d Barrels per day

BCeF Biomass conversion and expansion factor

BeF Biomass expansion factor

BNF Biologicalnitrogenfixing

BoD Biological oxygen demand

C Carbon

C2F6 Carbonhexafluoroethane

CF4 Carbontetrafluoromethane

CFC Chlorofluorocarbons

CH4 Methane

Co Carbon monoxide

Co2 Carbon dioxide

Co2eq Carbon dioxide equivalent

CrF Common reporting format

DaFF Department of agriculture affairs, Forestry and Fisheries

Dea Department of environmental affairs

Deat Department of environmental affairs and tourism

DFiD Department for international Development

DM Dry matter

DMD Dry matter digestibility

DMr Department of Mineral resources

DMe Department of Minerals and energy

Doe Department of energy

DoM Dead organic matter

Dti Department of trade and industry

DWa Department of Water affairs

LIST OF AbbREVIATIONS


DWaF Department of Water affairs and Forestry

eF emission factor

F-gases Fluorinated gases: e.g., HFC, pFC, SF6 and NF3

FoD First order decay

FolU Forestry and other land Use

FSa Forestry South africa

GDp Gross domestic product

Gei Gross energy intake

GFrSa Global Forest resource assessment for South africa

Gg Gigagram

GHG Greenhouse gas

GHGi Greenhouse Gas inventory

GiS Geographical information Systems

GpG Good practice Guidance

GWH Gigawatt hour

GWp Global warming potential

HFC Hydrofluorocarbons

HWp Harvested wood products

ieF implied emission factor

iNC initial National Communication

ipCC intergovernmental panel on Climate Change

ippU industrial processes and product Use

iSo international organization for Standardization

LPG Liquefiedpetroleumgas

lto landing/take off

MCF Methane conversion factor

MeF Manure emission factor

MW Megawatt

MWH Megawatt hours

MWtp Municipal wastewater treatment plant

list of abbreviations


NaeiS National atmospheric emissions inventory System

N2o Nitrous oxide

NCCC National Climate Change Committee

Ne Not estimated

NerSa National energy regulator of South africa

NGHGiS National Greenhouse Gas inventory System

Nir National inventory report

NiU National inventory Unit

NMvoC Non-methane volatile organic compound

Nox oxides of nitrogen

NtCSa National terrestrial Carbon Sinks assessment

NWBir National Waste Baseline information report

PFC Perfluorocarbons

ppM parts per million

prp pastures, rangelands and paddocks

Qa/QC Quality assurance/quality control

rSa republic of South africa

SaaQiS South african air Quality information System

SaiSa South african iron and Steel institute

SaMi South african Minerals industry

Sapia South african petroleum industry association

Sar Second assessment report

SaSQF South african Statistical Quality assurance Framework

SaDC Southern african Development Community

SF6 Sulphurhexafluoride

SNe Single National entity

taM typical animal mass

tar third assessment report (ipCC)

tJ terajoule

tM tier method


tMr total mixed ratio

toW total organics in wastewater

UN United Nations

UNep United Nations environmental programme

UNFCCC United Nations Framework Convention on Climate Change

Wri World resources institute

WWtp Wastewater treatment plant

vS volatile solids

list of abbreviations

GHG National inventory report for South africa 2000 -201224


UNITS, FACTORS AND AbbREVIATIONS

Multiplication factor Abbreviation Prefix Symbol

1 000 000 000 000 000 1015 peta p

1 000 000 000 000 1012 tera t

1 000 000 000 109 Giga G

1 000 000 106 Mega M

1 000 103 Kilo K

100 102 Hector H

0,1 10-1 Deci D

0,01 10-2 Centi C

0,001 10-3 Milli M

0,000, 001 10-6 Micro µ

Unit Equivalency

1 tonne (t) 1 Megagram (Mg)

1 Kilotonne 1 Gigagram (Gg)

1 Megatonne 1 teragram (tg)


ExECUTIVE SUMMARY

ES1 background information on South Africa’s GHG inventories

in august 1997 the republic of South africa joined the majority of countries in the international community in ratifyingtheUNFCCC.ThefirstnationalGHGinventoryin South africa was prepared in 1998, using 1990 data. it was updated to include 1994 data and published in 2004. it was developed using the 1996 ipCC Guidelines for National Greenhouse Gas inventories. For the 2000 national inventory, a decision was made to use the recently published 2006 ipCC Guidelines to enhance accuracy and transparency, and also to familiarise researchers with the latest inventory preparation guidelines. Following these guidelines, in 2014 the GHG inventory for the years 2000 to 2010 were compiled.

this report documents South africa’s submission of its national greenhouse gas inventory for the year 2012. it also reports on the GHG trends for the period 2000 to 2012. it is in accordance with the guidelines provided by the UNFCCC and follows the 2006 ipCC Guidelines and ipCC Good practice Guidance (GpG). the Common ReportingFormat(CRF)spreadsheetfileswereusedinthe compilation of this inventory. this report provides an explanation of the methods (tier 1 and tier 2 approaches), activity data and emission factors used to develop the inventory. in addition, it assesses the uncertainty and describes the quality assurance and quality control (Qa/QC) activities. Quality assurance for this GHG inventory was undertaken by independent reviewers.

Development of the National GHG Inventory System (NGHGIS)

During the compilation of the 2010 inventory there were several challenges that affected the accuracy and completeness of the inventory, such as application of lower tier methods as a result of the unavailability

of disaggregated activity data, lack of well-defined institutional arrangements, and absence of legal and formal procedures for the compilation of GHG emission inventories. South africa is currently in the process of creating a national GHG inventory system that will manage and simplify its climate change obligations to the UNFCCC process. this system will ensure: a) the sustainability of the inventory preparation in the country, b) consistency of reported emissions and c) the standard quality of results. the NGHGiS will ensure that the country prepares and manages data collection and analysis, as well as all relevant information related to climate change in the most consistent, transparent and accurate manner for both internal and external reporting. reliable GHG emission inventories are essential for the following reasons:

• Tofulfiltheinternationalreportingrequirementssuch as the National Communications and Biennial Update reports;

• to evaluate mitigation options;

• to assess the effectiveness of policies and mitigation measures;

• to develop long term emission projections; and

• to monitor and evaluate the performance of South africa in the reduction of GHG emissions.

South africa is developing a National atmospheric emissions inventory System (NaeiS) that will manage the mandatory reporting of GHG emissions. Due to their complex emission estimating methods, emission sectors such as agriculture, forestry and land use, and waste are to be estimated outside the NaeiS. the NaeiS, in turn, will ingest the outputs of models used in these sectors so that itcangenerateanationalemissionsprofile(Figure A).

the successful implementation of such an information management system is highly reliant on the development

executive Summary


of the NGHGiS which covers the GHG emissions inventory compilation process. the NGHGiS will be managed by a group of experts known as the National inventory Unit (NiU). the NiU will include sector experts, quality assurance and control specialists, an inventory co-ordinator, etc. the unit will be managed through the Dea.

the NGHGiS will include:

• the formalization of a National entity (the Dea) responsible for the preparation, planning, management , review, implementat ion and improvement of the inventory;

• legal and collaborative arrangements between

the National entity and the institutions that are custodians of key source data;

• a process and plan for implementing quality assurance and quality control procedures;

• the alignment of the NGHGiS with the South african Standard Quality assessment Framework;

• a process to ensure that the national inventory meets the standard inventory data quality indicators of accuracy, transparency, completeness, consistency and comparability; and

• a process for continual improvement of the national inventory.

FigureA: InformationflowinNAEIS.

Emissionsallocation

Emissionsreporting

Emissionsreporting

Input data preparation for

allocation

Processor

ActivityData

EF

Qa/QC doneoutside SaaQiS


Current inventory process

in the 1990, 1994 and 2000 GHG inventories for South africa activity and emission factor data were reported in the ipCC worksheets and the reports were compiled from this data. Supporting data and methodological details were not recorded, which made updating the inventory averydifficultandlengthyprocess.Inthe2000–2010GHG inventory more emphasis was placed on building up the annual data sheets and creating improved trend information. this led to better data records, but still very little supporting data and method details were kept. also, in all previous inventories the quality control procedures and uncertainty estimates were limited. as South africa moves forward, more emphasis has been placed on improving the documentation of inventory data and documents, as well as on uncertainty and quality control to improve the transparency of the inventory. the 2012 inventory has come a long way in addressing some of these issues.

the stages of the inventory update and improvement process (Figure b) were:

• Data collection, which involved:

o assessing data requirements;

o sourcing data from various stakeholders;

o screening data and selecting the appropriate data sets; and

o data quality control.

• Data input into the inventory database.

• Metadata input into the inventory database, which included:

o documentation of people responsible for inputting the data;

o specificdatacalculations;

o data sources;

o recalculations undertaken; and

o linkstodataorreferencefiles.

• Uncertainty analysis.

• Compilation of the inventory report.

• Quality control and quality assurance involving:

o assessment of the quality of the inventory;

o routine and consistent checks to ensure data integrity, correctness and completeness;

o identificationoferrorsandomissionsbothinthecalculations and report compilation;

o a public commenting process; and

o an external review.

Institutional arrangements for inventory preparation

the Dea is responsible for the co-ordination and management of all climate change-related information, including mitigation, adaption, monitoring and evaluation, and GHG inventories. although the Dea takes a lead role in the compilation, implementation and reporting of the national GHG inventories, other relevant agencies and ministries play supportive roles in terms of data provision across relevant sectors. Figure C gives an overview of the institutional arrangements for the compilation of the 2000 – 2012 GHG emissions inventory.

Organisation of report

this report follows a standard Nir format in line with the UNFCCC reporting Guidelines. Chapter 1 is the introductory chapter which contains background information for South africa, the country’s inventory preparation and reporting process, key categories, a description of the methodologies, activity data and emission factors, and a description of the Qa/QC process. a summary of the aggregated GHG trends by gas and emission source is provided in Chapter 2. Chapters 3 to 6 deal with detailed explanations of the emissions in

executive Summary


Figu

re B

: o

verv

iew

of t

he p

hase

s of

the

GH

G in

vent

ory

com

pila

tion

and

impr

ovem

ent

proc

ess.

Dat

a C

olle

ctio

n

Dat

a C

olle

ctio

n

Stakeholders

ener

gy

ippU

aFo

lU

Was

te

QC

5

Dat

a C

olle

ctio

n

trend and overall summary dataen

ergy

ippU

aFo

lU

Was

te

Dat

a C

olle

ctio

n

ener

gy

ippU

aFo

lU

Was

te

Dat

a C

olle

ctio

n

ener

gy

ippU

aFo

lU

Was

te

Key category and uncertainty analysis

Dat

a In

put

Inventory database

Dat

aO

utpu

t

GHG Inventory report

Final draft GHG inventory report

inventory report compilation

Final GHG inventory

QC

/QA

Dat

abas

e

QA

QC

1

QC

3

QC

2

QC

4


Figure C: institutional arrangements for the compilation of the 2000 – 2012 inventories.

the energy, ippU, aFolU and waste sectors, respectively. they include an overall trend assessment, methodology, data sources, recalculations, uncertainty and time-series

consistency, Qa/QC and planned improvements and recommendations.

DEA(National entity)

Chief Directorate: Monitoring and Evaluation (GHG inventory Unit)

Energy IPPU AFOLU Waste

DoE

DMR

Eskom

Coaltech

SAPIA

Chambersof Mines

NERSA

DoT

DoE

DMR

Eskom

Coaltech

SAPIA

Chambersof Mines

NERSA

DoT

GTI

ARC

FSA

CSIR

AgriSA

SAPA

UP/TUT NWU

DAFF

StatsSA

DEA-GIS unit

CSIR

DEA Waste Unit

StatsSA

executive Summary


ES2 Summary of South Africa’s GHG emissions

Overview of national emission and removal trends

in 2012 the total GHG emissions in South africa were estimated to be at 539 112 GgCo2eq (excl. FolU) (Table A). emissions (excl. FolU) increased by 21.7% between 2000 and 2012. there were decreases of 0.7% between 2004 and 2005 and 1.6% between 2007 and 2008, with the highest decrease of 2.7% between 2010 and 2011. including FolU, which was estimated to be a net carbon sink, the total GHG emissions in 2012 were reduced to 518 297 Gg Co2eq (Table A). emissions including FolU showed an increase of 19.3% over the 12 years from 434 304 Gg Co2eq in 2000. the declines in 2005 (0.7 %), 2008 (0.5 %), 2009 (0.2%) and 2011 (2.9 %) were still evident.

Overview of source and sink category emission estimates and trends

emissions from the energy sector increased by 25.0% between 2000 and 2012, while the ippU and waste sectors increased by 10.6% and 78.5%, respectively, over the same period (Table b). aFolU emissions declined by 32.1%.

the energy sector contributed 77.3% to the total GHG inventory (excl. FolU) in 2000 and this increased to 79.4% in 2012 (Figure D). there was a general increase in the

energy sector contribution; however slight declines were evident in 2001, 2008 and 2011. the aFolU sector (excl. FolU) is the second-largest contributor, followed by the ippU sector. aFolU (excl. FolU) contributed 12.3% in 2000, which declined to 9.6% in 2012, while in 2000

table a: trends in the national GHG emissions, including and excluding FolU, between 2000 and 2012.

Total (excl. FolU)

Total (incl. FolU)

(Gg Co2 eq)

2000 443 133 434 304

2001 442 966 434 093

2002 456 953 445 545

2003 478 546 467 765

2004 497 630 488 744

2005 494 270 485 487

2006 502 897 491 708

2007 529 335 517 168

2008 520 834 514 763

2009 526 633 513 509

2010 544 758 529 391

2011 529 969 514 257

2012 539 112 518 297

TableB: TrendsandlevelsinGHGemissionsfor2000and2012classifiedbysector

Total (excl. FolU) Total (incl. FolU) Percentage change between 2000 and 2012(Gg Co2 eq)

energy 342 592 428 112 25.0%

ippU 33 563 37 129 10.6%

aFolU 45 860 31 128 -32.1%

Waste 12 288 21 928 78.5%

Total (incl. FOLU) 434 304 518 297 19.3%


Figure D2: GHG emissions for South africa between 2000 and 2012 by sector with FolU included.

Figure D1: GHG emissions for South africa between 2000 and 2012 by sector excluding FolU.

ippU contributed 7.6%, which declined to 6.9% in 2012. the ippU sector showed declines in annual emissions in 2003, 2007 to 2009 and 2012. the aFolU sector (excl. FolU) showed a general decline with small annual increases (of less than 5%) in 2002, 2006, 2008, 2010 and 2011. the waste sector showed a steady increase in its contribution from 2.7% in 2000 to 4.0% in 2012. including the FolU component in the aFolU sector decreased

the total GHG emissions to 518 297 Gg Co2eq in 2012. this also changed the contribution from the energy, ippU, aFolU and waste sectors to 82.6%, 7.2%, 6.0% and 4.2%, respectively, in 2012 (Figure D1 and D2).

the top-emitting categories in 2000 and 2012 (excl. FolU) are shown in Figure E.

executive Summary


Figure e: the highest emitting categories in the South african GHG inventory (excl. FolU) in 2000 (top) and 2012 (bottom).

Figure F: Gas contribution to South africa’s GHG emissions between 2000 and 2012.

1.A.5.-Non-Specified2.B - Chemical industry

4.D - Wastewater treatment and Discharge2.a - Mineral industry

4.a - Solid Waste Disposal1.a.4 - other Sectors

2.C - Metal industry

3.a - livestock1.a.2 - Manufacturing industries and Construction

1.B - Fugitive emissions from fuels1.a.3 - transport

1.a.1 - energy industries

2000

GHG eMiSSioNS (exClUDiNG FolU) iN MtCo2-eQ0.00 50.00 100.00 150.00 200.00 250.00

1.A.5.-Non-Specified2.B - Chemical industry

4.D - Wastewater treatment and Discharge2.a - Mineral industry

4.a - Solid Waste Disposal

1.a.2 - Manufacturing industries and Construction3.a - livestock

2.C - Metal industry1.B - Fugitive emissions from fuels

1.a.4 - other Sectors1.a.3 - transport

1.a.1 - energy industries

2012

GHG eMiSSioNS (exClUDiNG FolU) iN MtCo2-eQ0.00 50.00 100.00 150.00 200.00 250.00 300.00


Overview of gas emission estimates and trends

South africa’s GHG emissions in 2012 were dominated by Co2 (83.7 %), followed by CH4 (10.7%) and N2o (4.9%). F-gases contributed less than 1% (Figure F). Co2 and CH4 emissions excluding FolU (incl. FolU) increased by 23.9% (21.2 %) and 15.1% (14.6 %), respectively, between 2000 and 2012, while N2o emissions showed a decline of 5.6%. pFC emissions doubled over the 12-year period, while HFC’s (only included from 2005) increased by 65.8% between 2005 and 2012.

Key categories

Key categories refer to the emission sources that contribute about 95% of the total GHG emissions in the country.Thekeycategorieswereidentifiedbycarryingoutthe ipCC tier 1 level and trend assessment with the 2000 and 2012 GHG inventories. the level assessment showed the key categories for 2012 (excluding FolU1) to be energy industries (solid fuels) – main activity electricity and heat, road transportation (liquid fuels), energy industries (liquid fuels) – manufacture of solid fuels and other energy industries, manufacturing industries and construction (solid fuels), and enteric fermentation - cattle; while the trend assessment (base year being 2000) indicated the key categories were other sectors (solid fuels) – residential, energy industries (solid fuels) – main activity electricity and heat production, manufacturing industries and construction (solid fuels), energy industries (liquid fuels) – manufacture of solid fuels and other energy industries, and enteric fermentation – cattle. if the land sector is included then land converted to forest landbecomesthefourthandfirstkeycategoriesinthelevel and trend assessments, respectively. Further details are provided in Section 1.3.4 with full results for 2012 given in Appendix b.

ES3 South Africa’s indicators

Figure G presents South africa’s emissions intensity between 2000 and 2012. the carbon intensity of the economy has dropped steadily over the decade. this is largelyduetogrowthintheservicesandfinancialsectors,a decline in the manufacturing sector and stagnation in the mining sector. the global economic crisis has had an impact on the carbon-intensity of the national energy supply even though there is generally stagnation elsewhere in the time series due to an unchanged energy supply mix.

ES4 Other information

General uncertainty evaluation

Uncertainty analysis is regarded by the ippC Guidelines as an essential element of any complete inventory. Chapter 3 of the 2006 ipCC Guidelines describes the methodology for estimating and reporting uncertainties associated with annual estimates of emissions and removals. there are two methods for determining uncertainty:

• tier 1 methodology which combines the uncertainties in activity rates and emission factors for each source category and GHG in a simple way; and

• tier 2 methodology which is generally the same as tier 1; however, it is taken a step further by considering the distribution function for each uncertainty, and then carries out an aggregation using the Monte Carlo simulation.

the reporting of uncertainties requires a complete understanding of the processes of compiling the inventory,

1 FolU is the forestry and other land use component of aFolU. it includes the land and the harvested wood products sub-categories of the aFolU sector.

executive Summary


sothatpotentialsourcesofinaccuracycanbequalifiedandpossiblyquantified.The2010inventory(Departmentof environmental affairs, 2014) did not incorporate an overall uncertainty assessment due to a lack of quantitative and qualitative uncertainty data. in this inventory there has been an attempt to incorporate an overall uncertainty assessment through the utilization of the ipCC uncertainty spread sheet. a trend uncertainty between the base year and 2012, as well as a combined uncertainty of activity data and emission factor uncertainty was determined using an approach 1. this inventory includes uncertainty assessment for the energy and ippU sectors only, but the other sectors will be included in the next inventory. the total uncertainty for the energy sector was determined to be 6.8%, with a trend uncertainty of 6.2%. the ippU sector has an uncertainty of 30.8% (8.83% excluding section 2.F).

Quality control and quality assurance

in accordance with ipCC requirements, the national GHG inventory preparation process must include quality control and quality assurance (QC/Qa) procedures. the objective of quality checking is to improve the transparency, consistency, comparability, completeness, and accuracy of the national greenhouse gas inventory. QC procedures, performed by the compilers, were carried out at various stages throughout the inventory compilation process (Figure b). Quality checks were completed at four different levels, namely (a) inventory data (activity data, eF data, uncertainty, and recalculations), (b) database (data transcriptions and aggregations), (c) metadata (documentation of data, experts and supporting data), and (d) inventory report. Quality assurance was completed through a public review process as well as an independent review.Theinventorywasfinalizedonceallcommentsfrom the quality assurance process were addressed.

Figure G: South africa’s emissions intensity between 2000 and 2012 (Department of environmental affairs, 2014)

Carbon intensity - energy Supply (Co2eq/tpeS):% Change (2000-2012) = (-) 8.7%(a) Carbon intensity or our energy supply impacted by the economic crisis;(b) Flactuations in the time series also influencedbytheenergycrisis.

energy intensity - economy (tpeS/GDp):% Change (2000-2012) = (-) 13%(a) More economic value derived per unit of energy supplied to the economy.

Carbon intensity - economy (Co2eq/GDp):% Change (2000-2012) = (-) 21%economy becoming greener thanks to:(a)Significantgrowthintheservicesandfinancialsectors;(b) reduction and stagnation in the manufacturing and mining sectors.

Greenhouse Gas Emissions - Indicators

2000 2002 2004 2006 2008 2010 2012

indi

cato

r Sc

ale

4.5

4

3.5

3

2.5

2

Co2eq/tpeS (tonnes/toe) Co2eq/GDp (tonnes/’100 r) tpeS/GDp (GJ/’thou r)


table C: activities in the 2012 inventory which were not estimated (Ne), included elsewhere (ie) or not occurring (No).

NE, IE or NO

Activity Comments

NE

Co2 and CH4 fugitive emissions from oil and natural gas operations

emissions from this source category will be included in the next inventory submission covering the period 2000-2014

Co2, CH4 and N2o from spontaneous com-bustion of coal seams

New research work on sources of emissions from this category will be used to report in the next emission inventory submission

CH4 emissions from abandoned mines New research work on sources of emissions from this category will be used to report emissions in the next inventory submission

other process use of carbonates

electronics industry a study was to be undertaken in 2015 to understand emissions from this source category

indirect N2o emission due to nitrogen deposition

HWp from solid waste this will be included in the next inventory

Co2, CH4 and N2o emissions from com-bined heat and power (CHp) combustion

systems

CH4, N2o emissions from biological treat-ment of waste

Co2, CH4 and N2o emissions from incinera-tion and open burning of waste

DoM emissions from croplands, grasslands, settlements and other lands

IE

Co2, CH4 and N2o emissions from wa-ter-borne navigation

Fuel consumption for this source-category is included elsewhere. Accuratequantificationoffuelconsumptionattributedtowa-

ter-borne navigation was to be undertaken in 2015 and a progress report will be included in the next inventory submission

Co2, CH4 and N2o emissions from off-road vehicles and other machinery

Fuel use associated with this source category is included in road transportation.AnewstudythatwastobefinalisedinOctober2015 on fuel apportionment to all demand sectors will help with

accurate allocation of fuels.

ozone depleting substance replacements for fireprotectionandaerosols

emissions from these sources are assumed to be accounted for in theTier1methodologyforquantificationofHFCemissions

NO

other product manufacture and use

rice cultivation this activity does not occur in South africa

Co2, CH4 and N2o emissions from Soda ash production

this activity does not occur in South africa

Co2 from Carbon Capture and Storage this activity does not occur in South africa

Co2, CH4 and N2o emissions from adipic acid production


Co2, CH4 and N2o Caprolactam, Glyoxal and Glyoxylic acid production


precursor emissions have only been estimat-ed for biomass burning, and only for Co and

Nox

precursor emissions from all sources will be included in the next inventory

executive Summary


Completeness of the national inventory

the GHG emission inventory for South africa is not complete, mainly due to data unavailability, and lack of data and estimation methodologies. Table C describes theactivitiesidentifiedbytheIPCC2006Guidelinesthatwere not estimated (Ne), included elsewhere (ie) or not occurring (No).

Methodological changes, recalculations and improvements

in the past year various improvements have been made to the GHG inventory due to the incorporation of more detailed activity data, updated emission factors and more consistent land-cover maps. in the energy sector recalculation were required as activity data and emission factors were improved. activity data was improved for energy industries (data obtained directly from national utility), transport (increased the use of Sapia data) and other sectors (inclusion of SaMi data). Further, for coal countryspecificemissionfactorswereappliedandtheN2o emission factor was corrected. recalculations were complete for all years between 2000 and 2012 and the changes lead to a slight increase (1.5%) in energy emissions (Table D).

the ippU emissions were recalculated due to updates in the iron and steel production emission factors, updated ferromanganese activity data, updated oDS and zinc production activity data. the recalculation meant a 20% reduction in the ippU emissions in 2010 (Table D), mostly because of the emission factor updates in the iron and steel production calculations.

SignificantupdatesweremadeintheAFOLUsector.Inthe agriculture sector small corrections were made to beef cattle and poultry population data, and emissions from game on privately owned parks was included. in the section on land use, new land-cover maps with a higher resolution were introduced, and emissions from DoM in forest lands and emissions from other lands were included. Updates to some of the biomass data were also made to incorporate data from the recent National terrestrial

Carbon Sinks report (Department of environmental affairs, 2015). the soil carbon methodology was updated to incorporate soil types, while the HWp data was updated based on the new Kp Supplementary Gp guidelines.

For the purpose of this report, the GHG emissions for the period 2000 to 2012 were recalculated using the updated activity data and emission factors so as to form a more consistent time series. in this way, the trends over time can be assessed. all the updates and recalculation methods and procedures are discussed in detail in Chapters 3 to 6 of this report.

the recalculations for the aFolU sector had the largest impact on the total emissions (Table D) with a 32.5% and 29.2% increase on the sector estimates in 2000 and 2010, respectively. the majority of this change was due to the corrected HWp estimates (this falls under the ipCC category 3D other) and the updated land change maps. the increases in the aFolU sector impacted the national total by 2.2% and 1.4% in 2000 and 2010, respectively (Table D). Small adjustments and recalculations were made in the energy, ippU and waste sectors, which contributed 0.7%, 0.3% and 0.1% to the increase in the national total, respectively.

The Greenhouse Gas Improvement Programme (GHGIP)

the main challenge in the compilation of South africa’s GHG inventory remains the availability of accurate activity data. the Dea is in the process of implementing a project that will ensure easy accessibility of activity data. it has initiated a new programme called the National Greenhouse Gas improvement programme (GHGip), whichcomprisesaseriesofsector-specificprojectsthatare targeting improvements in activity data, country-specificmethodologiesandemissionfactorsusedinthemostsignificantsectors.Table E and Table F summarize some of the projects that are under implementation as part of the GHGip.


table D: recalculated sector GHG emission estimates for 2000 and 2010 for South africa.

Initial emissions

Recalculated emissions % change

% impact on national total (incl. FOLU)(Gg CO2 eq)

energy2000 337 381 342 592 1.5% 1.2%

2010 428 368 435 117 1.6% 1.6%

ippU2000 44 907 33 563 -25.3% 2.6%

2010 44 350 35 463 -20.0% 2.0%

total aFolU2000 30 496 45 860 50.3% 3.5%

2010 25 713 38 456 49.6% 2.9%

3a livestock 2000 31 118 31 162 0.1% 0.01%

2010 28 986 29 412 1.5% 0.1%

3B land2000 -18 492 -8 520 53.9% 2.3%

2010 -33 224 -14 870 55.2% 4.2%

3C aggregated sources and non-Co2

emissions on land

2000 23 656 23 526 -0.5% -0.0%

2010 22 802 23 577 -3.4% -0.2%

3D other2000 -5 785 -312 -94.6% 1.3%

2010 -6 204 -491 -92.1% 1.3%

Waste2000 12 433 12 288 -1.2% -0.03%

2010 19 806 20 354 2.8% 0.13%

executive Summary


table e: Dea driven GHGip projects

Sector baselineNature of

methodological improvement

Partner Completion date

power generation [implications for other sectors]

Using ipCC default emission factors

Development of country-specificCO2 emission factors for

the Stationary

eskom, Sasol, GiZ october 2015

iron & steel

Using a combination of ipCC default

factors and assumptions based onmaterialflows

Combustion of Fuels in the electricity

Mittal Steel, SaiSi December 2015

transport sector [implications for other sectors]


Generation Sector Dot September 2017

Coal-to-liquids (Ctl)

allocation of emissions not

transparently done, not accounting for

all emissions

Shift towards a material balance

approachSasol December 2015

Ferro-alloy production

Using a combination of ipCC default

factors and assumptions based onmaterialflows

Development of country-specific

Co2, CH4, and N2o emission factors for road transportation.

xstrata, Ferro-alloy producers'

associationMay 2017

aluminium production


tier 2 approach BHp Billiton December 2014

Petroleumrefining

Not accounting for all emission sources.

Data time series inconsistencies

Shift towards a material balance

approach

Sapia in collaboration with all

refineriesDecember 2015


table F: Donor funded GHGip projects

Project Partner Objective Outcome Timelines

Development of a formal GHG

National inventory System

Norwegian Ministry of Foreign affairs

Helping South africa develop its national

system

Sa GHG inventories are documented and man-

aged centrally2015-2017

Stationary combustion eFs

GiZ,eskom, Sasol (power

utility)

to develop emissions factors for stationary combustion using the

power generation sector as a pilot

emissions from key sectors based on coun-try-specificinformation

2014-2015

land-cover mapping DFiD-UKto develop land-use

maps for 2-time steps [1990, 2013/14]

land-use change matrix developed for 36 ipCC land-use classes to de-

tect changes

2014-2015

aFolU sector activity data

improvementsDFiD-UK

to improve the activity data for agriculture,

croplands and land-use change

improved emission estimates for the aFolU

sector2014-2015

Waste-sector activity data

improvement project

african Development Bank

(afDB)

to improve waste-sector GHG emissions esti-

mates and address data gaps

Waste-sector GHG inventory is complete,

accurateandreflectiveofnational circumstances

2016-2017

Compliance with the SaSQaF

Statistics South africa

align GHG national inventory system with the SaSQaF to ensure quality of the inventory

the national GHG inventory and its

compilation processes endorsed through the SaSQaF evaluation

2014-2016

Survey on HFC consumption,

production and application.

GeF

to collect information on main application areas

for HFCs and pFCs as oDS substitutes

this will improve the estimation accuracy

of emissions from this sector

2016-2016

road transport Modelling Study

GeF

to quantify information on vehicle Kilometres travelled (vKt) in the

transport sector

this will improve the estimation accuracy

of emissions from this sector

2016-2017

executive Summary


ES5 Conclusions and recommendations

the 2000 to 2012 GHG emissions results revealed an increase in emissions from the energy and waste sectors, and an increase in the sink for the aFolU agriculture subsector. ippU emissions increased and declined again over the 12 years. the compilation of the GHG inventory continues to be a challenge, especially in the availability of activity data for the computation of GHG emissions. the inclusion of the land subsector in the aFolU sector caused a greater annual variation in the aFolU emission numbers, but there was a general increase in the sink capacity of the aFolU sector.

the energy sector in South africa continued to be the main contributor of GHG emissions (>75%) and was found to be a key category each year, therefore it is important that activity data from this sector is always available to ensure that the results are accurate. the accurate reporting of GHG emissions in this sector is also important for mitigation purposes.

the ippU emission estimates are largely derived from publicly available data from public institutions and sector-specificassociations.Sourcingofinformationatthe company level will enhance the accuracy of emission estimates and help reduce uncertainty associated with the estimates.

the aFolU sector was also highlighted as an important sector as it (excl. FolU) has a contribution greater than the ippU sector, and enteric fermentation is one of the top-10 key categories each year. the land subsector was also an important component of this sector because it was estimated to be a sink. South africa continues to require a more complete picture of this subsector and itisrecommendedthatmorecountry-specificdataandcarbon modelling be incorporated to move towards a tier 2 or 3 approach. this subsector also has important mitigation options for the future, and understanding the sinks and sources will assist in determining its mitigation potential.

in the waste sector the emission estimates from both the solid waste and wastewater sources were largely computed using default values suggested in the ipCC 2006 Guidelines, which could lead to large margins of error for South africa. South africa needs to improve the data capture of the quantities of waste disposed intomanagedandunmanagedlandfills,aswellasupdatewaste composition information and the mapping of all the wastewater discharge pathways. this sector would also benefitfromtheinclusionofmoredetailedeconomicdata (e.g. annual growth) broken down by the different population groups. the assumption that GDp growth is evenly distributed across the different populations groups is highly misleading and exacerbates the margins of error.


1.1 Climate change and GHG inventories

the republic of South africa ratified the UNFCCC and is therefore required to undertake several projects related to climate change. these include the preparation of greenhouse gas inventories, one of the outputs for the National Communications to the UNFCCC.

SouthAfrica’sfirstnationalGHGinventorywascompiledin 1998 using activity data for 1990. the second national GHG inventory used 1994 data and was published in 2004. Both the 1990 and 1994 inventories were compiled based on the 1996 ipCC Guidelines.

the third national GHG inventory was compiled in 2009 using activity data from 2000. For that inventory the 2006 ipCC Guidelines were introduced. those guidelines ensured accuracy, transparency and consistency. the 2006 IPCCGuidelinesmadesignificantchanges,particularlyonthe restructuring of inventory sectors. Countries are not required by the UNFCCC to report their GHG emissions using the updated guidelines; however, it was decided tousethelatestIPCCGuidelinestoavoiddifficultiesconverting from the 1996 to the 2006 guidelines in the future. in 2014 South africa prepared its fourth national inventory, which included annual emission estimates for 2000 to 2010.

1.2 Country background

1.2.1 National circumstances

South africa is a culturally diverse developing country with11officiallanguages.Itisasignificantindustrialandeconomic power in africa and has the largest economy in southern africa. the country has well-developed mining, transport, energy, manufacturing, tourism, agriculture, commercial timber and pulp production, and

service sectors. it is a net exporter of energy, food, and telecommunications and other services to neighbouring countries. South africa shares borders with six countries: Namibia, Botswana, Zimbabwe and Mozambique, which lie to the north, and lesotho and Swaziland, which are landlocked within South africa.

Therearevariousfactorsthatcaninfluenceanation’sGHG emissions, including government’s infrastructure, population growth, geography, economic growth, energy consumption, technological development, climate and soils, agriculture, and land use management. South africa contributes to global climate change and, as a result, has taken steps to formulate measures to mitigate and adapt to a changing climate.

1.2.1.1 Government Structure

South africa is a multiparty democracy, with three tiers of government: national, provincial and local. responsibilities affecting economic development, energy and natural resources, among others, are shared across the three tiers.

1.2.1.2 PopulationProfile

South africa’s population in 2013 was 52.98 million (Statistics Sa, 2013). it had grown by 11.54% since 2001 to 2010 and the projected growth rate for 2010 to 2015 was 0.5% (Department of environmental affairs, 2011). international immigration is currently the main driving force of South africa’s population growth (Department of environmental affairs, 2014). Strong socio-economic and policy drivers of migration to urban centres have been at play, as indicated by the urban population increase from 52% to 62% over the past two decades (UNDp, 2010). South africa is one of the most urbanised countries in africa. there are nine provinces: eastern Cape, Free State, Gauteng, KwaZulu-Natal, limpopo, Mpumalanga, Northern Cape, North West and Western Cape.

1. introduction

1. INTRODUCTION


Gauteng is the most populated, with 24% of residents (Department of environmental affairs, 2014), followed by KwaZulu-Natal with 19.7%. the Northern Cape has the smallest population (2.2%) (Statistics Sa, 2013). South Africahas11officiallanguages,whichdisplaysitsculturaldiversity.

1.2.1.3 GeopgraphicProfile

South africa is a medium-sized country, which covers roughly 124 929 847 hectares of the southernmost part of the african continent. it measures approximately 1 600 km from north to south, and is about the same from east towest.Itliesbetween22°and35°south,flankedonthe west by the atlantic ocean and on the east by the indian ocean (GCSi, 2009). the coastline is more than 2 500 km long, extending southwards from the desert border with Namibia in the northwest to Cape agulhas, and then northwards to the indian ocean until it reaches the subtropical border with Mozambique in the northeast (GCSi, 2009). it has narrow coastal plateaus in the south and west which are edged by coastal mountain ranges. Further into the interior, an escarpment borders an extensive elevated plateau on which most of the central areas are 1 000 m above sea level. the main geographical features are the Drakensberg Mountains in the east, the Great escarpment in the northeast and the Great Karoo in the centre (Department of environmental affairs, 2014).

1.2.1.4 EconomicandIndustrialProfile

South africa is deemed a developing country with well-developed mining, transport, energy, manufacturing, tourism, agriculture, commercial timber and pulp production, and service sectors. the national GDp was $248 million in 2007, which translated to a per capita GDp of r36 461 (Department of environmental affairs, 2011). this increased to $380 billion (r2 835 087) in 2012 (Statistics Sa, 2013). Between 2004 and 2007 the GDp grew at an annual rate of 4.6% to 5.6%, but in 2008 it slowed to 3.8% (Table 1.1). that was due to the global economic recession. in November 2009 South africa showed that it had started to recover from the recession by achieving a growth of 0.9%. the hosting of the2010FIFAWorldCupcontributedsignificantlytothecountry’s economy, helping it to register a 3.4% growth in the fourth quarter of 2010 (Statistics Sa, 2013). in 2012, the agriculture, industry and service sectors accounted for 2.5%, 31.6% and 65.9% of GDp, respectively.

Annual inflationconsumerprice index lessmortgageinterest increased from 10% in March 2008 to 13.6% in august 2008. the annual rate of increase in food prices remained constant at 14.9% in March 2009, compared to a peak of 19.2% in august 2008. processed food prices remainedathighlevelsreflectinghighercostoftransport,wages and general production.

table 1.1: the GDp percentage growth in South africa between 2000 and 2012 (Statistics Sa, 2014).

Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

GDp % growth

4.2 2.7 3.7 2.9 4.6 5.3 5.6 5.4 3.2 -1.5 3.0 3.2 2.2


1. introduction

South africa has an abundant supply of mineral resources. its economy was originally built on natural resources, with agriculture and mining being the major components of the GDp. However, over the past few years thousands of jobs have been lost in the mining industry (Department of environmental affairs, 2011), nonetheless, South africa’s mining industry remains one of the country’s main employers.

the transport sector is dominated by road travel. South africa’s car ownership ratio is higher than the world average, which is attributed to the large distances between settlements and places of employment. Within the road transport sector 19% is due to private vehicle trips, and 11.5% due to minibus taxis. Sixty-three percent of the country’s commuters use minibus taxis, 22% use buses and 15% use trains (Department of environmental affairs, 2011). the minibus-taxi industry continues to grow annually.

With the growing population and economy, waste processing and disposal continue to be signif icant challenges. Households with adequate refuse removal services have remained at about 60% since 2006, but over 80% of these households are in urban areas. in rural areas, as little as 20% of households have access to these services.

1.2.1.5 NaturalResourcesProfile

South africa is located in a subtropical region, making it a warm and sunny country. its climate is moderated by oceanicinfluencesalongtheextensivecoastlineandbythe altitude of the interior plateau. its climate is generally warm temperate dry, however, there are exceptions, which create climatic diversity. there is a temperate Mediterranean-type climate in the southwest, a warm subtropical climate in the northeast, and a warm, dry desert environment in the central west and northwest. South africa is a semi-arid region with an east-west rainfall gradient ranging from less than 250 mm to over 1000 mm per year (State of environment report, 2006). Most

of the rainfall occurs in summer; however, the Western Cape receives most of its rainfall in winter (GCSi, 2009).

the relatively young geology in South african gives rise to high-nutrient soils. the Nama-Karoo biome of the central region comprises predominantly mudstones and sandstones of the Karoo Supergroup, while the Highveld grasslands are associated with high-nutrient soils of basalt and dolerite origin (palmer & ainslie, 2002). the Savannas of the Mpumalanga lowveld are associated with Gabbros and granites (sandy, moderate-nutrient status) of the Bushveld igneous Complex. the Cape Fold Mountains are siliceous rocks giving rise to litholic soils, while the lesotho Highlands are basaltic giving rise to mollisols (partridge, 1997).

the land cover in South africa is dominated by low shrublands (33%), grasslands (20%) and woodlands or open bush (10%) (Geoterraimage, 2015; see appendix G). Natural forests are very small in South africa, covering less than 0.5% of the land area. Settlements occupy approximately 2% of the land area (Geoterraimage, 2015). although 80% of South africa’s land surface area is used for agriculture and subsistence livelihoods, only about 11% has arable potential. the remaining 69% is used for grazing. areas of moderate to high arable potential occur mainly in the eastern part of the country, in Mpumalanga and Gauteng, with scattered patches in KwaZulu-Natal, the eastern Cape, and limpopo. the agricultural sector is very diverse, ranging from intensive, large-scale commercial agricultural to low-intensity, small-scale subsistence farming. roughly 11% of the land is formally cultivated, with a further 1.6% under subsistence farming. Maize and wheat are the dominant annual crops by area. plantations are based on non-native trees, with the dominant species being eucalyptus grandis.

land in South africa has enormous economic, social and environmental value, but a large proportion of the country’s land surface is susceptible to degradation. Schoeman et al. (2013) showed that there was a 1.2% increase in transformed land between 1994 and 2005.


Cultivation and degradation of natural cover contributed 10.5% and 4.5%, respectively, to this transformation. the rest was due to urban land use and forestry. Cultivated areas over this period declined from 12.4% to 11.9%.

1.2.1.6 Agriculture,ForestryandFisheriesProfile

Agriculture,forestryandfisheriestogetheraccountedfor less than 3% of GDp in 2006 (Department of environmental affairs, 2011). the agriculture sector is dominated, in economic output terms, by large-scale commercial farming, but there is a very important small-scale and subsistence sector. the total contribution of agriculture to the economy increased from r27 billion in 2001 to r36 billion in 2007. During the period 2008 to 2009 the sale of animals and animal products accounted for48.2%oftheincome,26.7%wasfromfieldcropsand25.1% was from horticulture (DSt, 2010). South africa’s largest agricultural commodity by mass in 2007 was sugar cane, followed by cattle meat, chicken meat, grapes and dairy.

South africa’s timber plantations are based on non-native trees and cover 1.3% of the land area (Geoterraimage, 2015). timber contributes more than r16 billion to South africa’s economy, with an annual production of 2.2 million m3 of commercial round wood (Department of environmental affairs, 2011). exports are mainly converted, value-added products, with raw material exports making up only 1.8% of the total. the main products exported are pulp and paper (73%), saw timber, wood chips and wattle extract.

Thecommercialandrecreationalfishingindustriesarearelatively small economic sector, contributing about 1% of GDp and valued at approximately r4 billion to 5 billion annually (Department of environmental affairs, 2011).

1.2.2 Institutional arrangements for inventory preparation

in South africa the Dea is the central co-ordinating and policy-making authority with respect to environmental conservation. the Dea is mandated by the air Quality act (act 39 of 2004) to formulate, co-ordinate and monitor national environmental information, policies, programmes and legislation. the work of the Dea is underpinned by the Constitution of the republic of South africa and all other relevant legislation and policies applicable to government to address environmental management, including climate change.

in its capacity as a lead climate institution, the Dea is responsible for co-ordination and management of all climate change-related information, such as mitigation, adaption, monitoring and evaluation programmes, including the compilation and update of GHG inventories. the branch responsible for the management and co-ordination of GHG inventories at the Dea is the Climate Change and air Quality Management branch, whose purpose is to improve air and atmospheric quality, as well as support, monitor and report international, national, provincial and local responses to climate change.

although the Dea takes a lead role in the compilation and reporting of the national GHG inventories, there are many other relevant agencies and ministries that play supportive roles in terms of data provision across relevant sectors. the major contributors are shown in Table 1.2 to 1.5.

Atthisstagethereisalackofwell-definedinstitutionalarrangements and an absence of legal and formal procedures for the compilation of GHG emission inventories. the structure and formalization of these institutional arrangements is currently being developed by the Dea so as to maintain a consistent and sustainable flowofdatainputfortheinventory.


table 1.2: institutional arrangements in the energy sector

Ministry / Agency Role

Department of energy (Doe)the Dea utilises national energy balances published by the Doe to estimate energy emissions.

Department of Mineral resources (DMr)

provides energy-related information in the form of the South african Minerals industry (SaMi) report.

eskomprovides information on GHG emissions associated with electricity genera-tion.

Coaltech provides information on fugitive emissions associated with solid fuels.

South african petroleum industry association (Sapia)

ProvidesinformationforestimatingGHGemissionsfrompetroleumrefining.

Chamber of Mines provides information for estimating GHG emissions from mining activities.

National energy regulator of South africa (NerSa)

provides information for estimating GHG emissions from independent power producers (ipps).

Department of transport provides information for estimating GHG emissions from the transport sector.

table 1.3: institutional arrangements in the ippU sector


Business Unity South africa (BUSa) and industry

provide GHG information for the ippU sector.

South african iron and Steel institute (SaiSi)

provides GHG information for the iron and steel industry.

association of Cementitious Material producers (aCMp)

provides GHG information for the cement industry.

Ferro-alloy producers' association (Fapa)

provides activity data on ferroalloy production.

Department of Mineral resources (DMr)

provides industrial processes-related information in the form of the South african Minerals industry (SaMi) reports.

Department of transport provides information for estimating GHG emissions from the transport sector.

1. introduction


table 1.4: institutional arrangements in the aFolU sector


Geoterraimage provides maps for the land sector.

agricultural research Council (arC)

provides data and technical support when compiling GHG emissions for agriculture.

Forestry South africa (FSa) provides data for the forestry sector.

agri South africa (agriSa)provides technical support for compiling GHG emissions from the agriculture sector.

South african poultry association (Sapa)

provides poultry population data.

CouncilforScientificandIndustrialresearch (CSir)

provides technical support when compiling GHG emissions for land.

University of pretoria (Up) and tshwane University of technology (tUt)

Undertake research related to agricultural emissions.

North-West University (NWU) provide support for the compilation of the aFolU sector inventory.

Department of agriculture, Forest-ry and Fisheries (DaFF)

provides data, statistics and information on the aFolU sector.

Statistics South africa (Stats Sa) provides statistical data for the aFolU sector.

Department of environmental affairs (Dea) Geographic information System (GiS) Unit

provides an archiving system for forestry maps.

table 1.5: institutional arrangements in the waste sector


CouncilforScientificandIndustrialresearch (CSir)

provides technical support when compiling GHG emissions for the waste sector.

Dea Waste Unit provides information on the waste sector.

Statistics South africa (StatsSa)provides statistical parameters that can be applied when estimating GHG emissions from the waste sector

Department of Water affairs (DWa)

provides information on the country’s wastewater treatment works.


1.3 Inventory preparation

it was decided before the preparation of the 2000 inventory that the 2006 ipCC Guidelines would be used to prepare South africa’s GHG inventory to ensure consistency, accuracy and transparency. to be consistent with previous inventories, this GHG inventory for 2012 was also prepared using the principles of the 2006 ipCC Guidelines. the method chosen for each source category is guided by the decision trees provided in the ipCC Guidelines. in general, the method selected depended on whether the source category was considered as the main category, and on the availability of data.

the collection of data and information is still a challenge when compiling the GHG inventory for South africa. the data and information are often collected from national aggregated levels rather than from point or direct sources. Thatmakestheuseofhigher-tiermethodsdifficult.Wheremore disaggregated data and emission factors were available, a higher-tier method was used to improve on the previous inventory. South africa’s aim is to incorporate morecountry-specificdataandmovetowardsaTier2or3 approach for the key categories in particular.

the Dea is in the process of implementing a data management system that will manage all aspects of the inventory compilation process. the NGHGiS will include:

• the formalization of a National entity (the Dea), which will include an inventory co-ordinator, sector-specificexpertsandaninventorycompiler.this group will be responsible for the preparation, planning, management, review, implementation and improvement of the inventory;

• legal and collaborative arrangements between the National entity and the institutions that are custodians of key source data and that will provide the required data for each sector;

• a process and plan for implementing Qa and QC procedures, including ensuring that the UNFCCC reporting and ipCC Guidelines have been properly

implemented;

• the alignment of the NGHGiS with the South african Standard Quality assessment Framework;

• a process to ensure that the national inventory meets the standard inventory data quality indicators of accuracy, transparency, completeness, consistency and comparability; and

• a process for continual improvements of the national inventory.

1.3.1 Data collection procedures and plans

a variety of data suppliers provide basic input data for emission estimates. Data collection and documentation were the responsibility of individual experts in each sector. Data came mostly from government institutions, local and international literature, and, to a lesser extent, from individual industrial plants and professional associations. Data access continues to be a challenge for South africa; therefore this inventory is not complete. Some sources or sinks have been omitted due to lack of appropriate data (see Section 1.6).

the Dea is in the process of developing a data collection plan that will improve the accessibility of activity data. the data collection procedures will include guidance on collection, collation and processing existing data, as well as providing guidance for the generation of new data. a data collection plan is required to meet data quality objectives regarding timeliness, consistency, completeness, comparability, accuracy and transparency. it is also good practice to periodically review the data collection process to maintain the quality and to guide progressiveandefficientinventoryimprovement.

the data collection plan will focus on providing guidance for three main aspects (Figure 1.1):

• Collection and collation of existing data:

o focus data collection on improving estimates for key categories which:

1. introduction


- are the largest,

- have the greatest potential to change, or

- have the greatest uncertainty;

o choose data collection procedures that iteratively improve the quality of the inventory in line with the data quality objectives; and

o collect data at a level of detail appropriate to the method used;

• Generation of new data:

o identifydatagapsandworktowardsfillingthesegaps; and

o put in place data collection activities that lead to the continual improvement of data sets used in the inventory;

• Maintenance of data supply and collection procedures:

o introduce agreements with data suppliers to support consistent and continuing information flows;and

o rev iew data col lect ion act iv i t ies and methodological needs on a regular basis.

1.3.2 Data archiving and storage

after the data has been through the data collection phase (collection, selection, generation, review and QC) it enters the GHG inventory database. in the past, each inventory was treated differently and separate worksheets were completed for each inventory. Furthermore, no documentation (of methodology or QC) or supporting data were recorded. the 2000 – 2010 inventory was the

Figure 1.1: GHG inventory data collection procedure

Existing data

Maintain data supply:• Collaboration• Workshops• information updates• tor’S• MoU’s

Data maintenance New data

Stak

ehol

ders

GHG Inventory input data

Review:• Method guidelines/protocols• Data sources• Data selection procedures• QC procedures

Obtain data:• Set up formal data requests

Determine data requirements

Identify data sources

Screen data sources

Identify data gaps

Fill data gaps

Time series:• Models • Splicing

Activitiy & EF data:• Measurement• Census/surveys• Models

Develop methodological guidelines or protocols

QC


Figure1.2:ConfigurationandfilescontainedintheGHGinventorydatabase.

Gen

eral

sec

tora

l dat

a

Ener

gy

GHG

Inve

ntor

y dat

a she

ets

GH

G M

etad

ata

GH

G In

vent

ory

QC

GH

G In

vent

ory

Repo

rt

IPPU

AFO

LU

Was

te

Qua

lity

Con

trol

_201

2En

ergy

IPPU

AFO

LU

Was

te

Com

bine

d re

port

:

•N

ir_2

010

•N

ir_2

012

Ener

gy_2

000

......

....

Ener

gy_2

012

Ener

gy_T

rend

s

Ener

gy_2

000

......

....

Ener

gy_2

012

Ener

gy_T

rend

s

AFO

LU_2

000

......

....

AFO

LU_2

012

HW

P ca

lcul

atio

n_20

12A

FOLU

_Tre

nds

Was

te_2

000

......

....

Was

te_2

012

IPCC

_Was

te_M

odel

_201

2W

aste

_Tre

nds

Unc

erta

inty

_201

2

Ener

gy_2

012_

Met

adat

aEn

ergy

_Sup

port

ing_

...

Ener

gy

IPPU

_201

2_M

etad

ata

IPPU

_Sup

port

ing_

...

IPPU

AFOL

U_201

2_Me

tadata

AFOL

U_Ag

ricult

ure_S

uppo

rting

_...AF

OLU_

Land

_Sup

porti

ng_...

AFO

LU

Was

te_2

012_

Met

adat

aW

aste

_Sup

port

ing_

...

Was

te

GH

G In

vent

ory

GH

G S

umm

ary_

2012

1. introduction


firstoneinwhichanannualtimeserieswasincluded,andbecause of the increase in the amount of data there was a focus on setting up a proper inventory database where all the data could be kept. the database will also assist in streamlining future GHG inventory updates. in this 2000 – 2012 inventory there has been a focus on documentation, transparency and quality control. additional spreadsheets for metadata and supporting documents, as well as quality controlanduncertaintyfileshavebeenaddedtothedatabase.

the database consists of a series of excel spreadsheets which are linked to ensure consistency and easy updating. the GHG inventory database contains five folders (Figure 1.2):

• General sectoral data;

• GHG inventory data sheets;

• GHG metadata;

• GHG inventory QC; and

• GHG inventory report.

1.3.3 brief description of methodologies and data sources

1.3.3.1 Methodologies

the 2006 ipCC Guidelines were used for estimating GHG emissions for South africa. there are four main inventory sectors, namely: energy, ippU, aFolU and waste. For the current inventory, data were gathered for the following gases: Co2, CH4, and N2o. Certain HFCs and pFCs were reported on in the ippU sector and Nox and Co were estimated for biomass burning emissions.

this section provides an overview of the methods used to estimate GHG emissions in South africa, and Table 1.6 summarizes the tier method and type of emission factors used in the calculations.

after data were collected, the sources quality assured, and the unit conversions completed, the GHG emissions were calculated by inventory experts using the following basic principle (ipCC 2006 Guidelines):

Emission = activity data x emission factor

as required by the 2006 ipCC Guidelines, the aFolU and waste sectors made use of more complex calculations and models which are described in Chapters 5 and 6.

When calculating emissions, the expert is responsible for quality assurance and checks, but the calculations are also checked by external parties to ensure accuracy and consistency. emission factors from national sources are the most accurate, but, where national emission factors are not available, default ipCC emission factors should be used. in most cases default factors where used where disaggregated data could not be obtained and the tier 1 approach was applied. More detailed methodologies for each sector and source category are presented in later chapters.

in January 2015 all countries will be mandated to use the 2006 ipCC Guidelines for the compilation and reporting of national GHG inventories under the UNFCCC reporting guidelines. in addition, non-annex i countries that are party to the convention shall use the GWp from the ipCC Second assessment report (Sar). in this current inventory, South africa has chosen to use GWps from the ipCC third assessment report (tar) (Table 1.7) so as to align with the 2006 ipCC Guidelines. Discussions are underway domestically to consider the possibility of using GWps from the ipCC Fourth assessment report (Far).


tabl

e 1.

6:

tie

r m

etho

d (t

M)

and

emis

sion

fact

or (

eF)

used

in t

his

inve

ntor

y in

the

est

imat

ion

of t

he e

mis

sion

s fr

om t

he v

ario

us s

ecto

rs.

GH

G S

ourc

e an

d Si

nk C

ateg

ory

CO

2C

H4

N2O

HFC

sP

FCs

SF6

tM

eFt

MeF

tM

eFt

MeF

tM

eFt

MeF

1.

Ene

rgy

a.

Fuel

com

bust

ion

1.

ener

gy in

dust

ries

t1, t3

t1,

CSa

t1, t1

DFb

t1, t1

DF

2.

Man

ufac

turi

ng in

dust

ries

and

con

stru

ctio

nt

1D

Ft

1D

Ft

1D

F

3.

tran

spor

tt

1D

Ft

1D

Ft

1D

F

4.

oth

er s

ecto

rst

1D

Ft

1D

Ft

1D

F

5.

Non-specified

t1

DF

t1

DF

t1

DF

B.

Fugi

tive

emis

sion

s fr

om fu

els

HFC

-152

at

2C

St

2C

S

t3

CS

t3

CS

t3

CS

2.

Indu

stri

al p

roce

sses

a.

Min

eral

pro

duct

st

1D

F

B.

Che

mic

al in

dust

ryt

1,

t3

DF

CS

t1

DF

t1,

t

3D

F

C.

Met

al in

dust

ryt

2,

t3

t2,

pS

ct

1D

Ft

2C

S

D.

Non

-ene

rgy

prod

ucts

from

fuel

s an

d so

lven

tst

1D

F

F. pr

oduc

t us

ed a

s su

bstit

utes

for

oD

St

1D

F

3.

AFO

LU

a.

live

stoc

k

1.

ente

ric

ferm

enta

tion

t2

CS

2.

Man

ure

man

agem

ent

t2

CS

t2

DF

B.

land

1.

Fore

st la

ndt

1,

t2

CS

DF

2.

Cro

plan

dt

1,

t2

CS

DF

3.

Gra

ssla

ndt

1C

SD

F

4.

Wet

land

st

1D

F

5.

Sett

lem

ents

t1

DF

6.

oth

er la

nds

t1

DF

C.

agg

rega

ted

sour

ces

and

non-

Co

2 em

issi

ons

1.

emis

sion

s fr

om b

iom

ass

burn

ing

t2

CS

t2

CS

t2

CS

2.

lim

ing

t1

DF

3.

Ure

a ap

plic

atio

nt

1D

F

4.

Dire

ct N

2o fr

om m

anag

ed s

oils

t1

DF

5.

indi

rect

N2o

from

man

aged

soi

lst

1D

F

6.

indi

rect

N2o

from

man

ure

man

agem

ent

t1

DF

D.

oth

er

7.

Har

vest

ed w

ood

prod

ucts

t1

CS

DF

4.

Was

te

a.

Solid

was

te d

ispo

sal

t1

DF

D.

Was

tew

ater

tre

atm

ent

and

disc

harg

et

1,

t2

CS

DF

t1

DF

aC

ountryspecific;

b D

efau

lt fa

ctor

s

cPlantSpecificemissionfactors

1. introduction


GH

G S

ourc

e an

d Si

nk C

ateg

ory

CO

2C

H4

N2O

HFC

sP

FCs

SF6

tM

eFt

MeF

tM

eFt

MeF

tM

eFt

MeF

1.

Ene

rgy

a.

Fuel

com

bust

ion

1.

ener

gy in

dust

ries

t1, t3

t1,

CSa

t1, t1

DFb

t1, t1

DF

2.

Man

ufac

turi

ng in

dust

ries

and

con

stru

ctio

nt

1D

Ft

1D

Ft

1D

F

3.

tran

spor

tt

1D

Ft

1D

Ft

1D

F

4.

oth

er s

ecto

rst

1D

Ft

1D

Ft

1D

F

5.

Non-specified

t1

DF

t1

DF

t1

DF

B.

Fugi

tive

emis

sion

s fr

om fu

els

HFC

-152

at

2C

St

2C

S

t3

CS

t3

CS

t3

CS

2.

Indu

stri

al p

roce

sses

a.

Min

eral

pro

duct

st

1D

F

B.

Che

mic

al in

dust

ryt

1,

t3

DF

CS

t1

DF

t1,

t

3D

F

C.

Met

al in

dust

ryt

2,

t3

t2,

pS

ct

1D

Ft

2C

S

D.

Non

-ene

rgy

prod

ucts

from

fuel

s an

d so

lven

tst

1D

F

F. pr

oduc

t us

ed a

s su

bstit

utes

for

oD

St

1D

F

3.

AFO

LU

a.

live

stoc

k

1.

ente

ric

ferm

enta

tion

t2

CS

2.

Man

ure

man

agem

ent

t2

CS

t2

DF

B.

land

1.

Fore

st la

ndt

1,

t2

CS

DF

2.

Cro

plan

dt

1,

t2

CS

DF

3.

Gra

ssla

ndt

1C

SD

F

4.

Wet

land

st

1D

F

5.

Sett

lem

ents

t1

DF

6.

oth

er la

nds

t1

DF

C.

agg

rega

ted

sour

ces

and

non-

Co

2 em

issi

ons

1.

emis

sion

s fr

om b

iom

ass

burn

ing

t2

CS

t2

CS

t2

CS

2.

lim

ing

t1

DF

3.

Ure

a ap

plic

atio

nt

1D

F

4.

Dire

ct N

2o fr

om m

anag

ed s

oils

t1

DF

5.

indi

rect

N2o

from

man

aged

soi

lst

1D

F

6.

indi

rect

N2o

from

man

ure

man

agem

ent

t1

DF

D.

oth

er

7.

Har

vest

ed w

ood

prod

ucts

t1

CS

DF

4.

Was

te

a.

Solid

was

te d

ispo

sal

t1

DF

D.

Was

tew

ater

tre

atm

ent

and

disc

harg

et

1,

t2

CS

DF

t1

DF

aC

ountryspecific;

b D

efau

lt fa

ctor

s

cPlantSpecificemissionfactors


table 1.7: Global warming potential (GWp) of greenhouse gases used in this report (Source: ipCC 2001).

Greenhouse gas Chemical formula TAR GWP

Carbon dioxide Co2 1

Methane CH4 23

Nitrous oxide N2o 296

Hydrofluorocarbons (HFCs)

HFC-23 CHF3 12 000

HFC-32 CH2F2 550

HFC-125 CHF2CF3 3 400

HF-134a CH2FCF3 1 300

HFC-143a CF3CH3 4 300

HFC-152a CH3CHF2 120

Perfluorocarbons (PFCs)

pFC-14 CF4 5 700

pF-116 C2F6 11 900

pFC-218 C3F8 8 600

pFC-31-10 C4F10 8 600

pFC-318 c-C4F8 10 000

pFC-4-1-12 C5F12 8 900

pFC-5-1-14 C6F14 9 000

Sulphur hexafluoride

Sulphurhexafluoride SF6 23 900

1. introduction


1.3.3.2 Data Sources

in general, the following primary data sources supplied the annual activity data used in the emission calculations:

• energy data:

o Department of Mineral resources (DMr);

o Department of energy (Doe);

o power utility plants;

o Chevron;

o SapreF;

o engen;

o energy balances and the periodically published Digests of South african energy Statistics provided an overview of the interrelations within South africa’s energy sector, by providing a breakdown of the different fuels and category sources;

o South african petroleum industry association (Sapia) provided other energy data, transport fuel data and crude oil production;

o petroSa;

o Sasol;

o eskom;

o Food and agriculture organization (Fao);

o Chamber of Mines provided information associated with GHG emissions from mining activities;

o Statistics South africa;

o National energy regulator of South africa (NerSa);

o Government Communications and information Systems (GCiS); and

o South african reserve Bank.

• industrial processes and product use:

o Business Unity South africa;

o Chemical and allied industries association (Caia);

o industry associations;

o South african Minerals industry (SaMi) report;

o Department of Mineral resources (DMr);

o Department of energy (Doe);

o South african iron and Steel institute (SaiSi);

o association of Cementitious Material producers (aCMp); and

o Direct communication with various industrial production plants.

• aFolU:

o Department of agriculture, Forestry and Fisheries (DaFF);

o Forest resource assessment (from DaFF and Fao);

o Department of environmental affairs (Dea);

o Forestry South africa (FSa);

o Geoterraimage (Gti);

o National terrestrial Carbon Sinks assessment (NtCSa) (Dea, 2015);

o Food and agriculture organization (Fao);

o agricultural research Council (arC);

o tshwane University of technology (tUt);

o University of pretoria (Up);

o North West University, potchefstroom (NWU);

o Statistics South africa;

o professional livestock associations and breed societies;

o MoDiS burnt area data.


• Waste:

o CentreforScientificandIndustrialResearch(CSir);

o Statistics Sa;

o World resources institute (Wri); and

o Dea waste baseline study.

1.3.4 brief description of key categories

Thekeycategoriesarethemostsignificantemissionsources in South africa. there are two approaches which can be used to determine the key categories; namely, the level approach and the trend approach. the former is used if only one year of data is available, while the latter can be used if there are two comparable years. the inventory provides emissions for more than one year; therefore both the level and trend assessments for key category analysis were performed.

the key categories have been assessed using the approach 1 level (l1) and approach 1 trend (t1) methodologies from the 2006 ipCC Guidelines (ipCC, 2006). the key categoryanalysisidentifieskeycategoriesofemissionsandremovals as those that sum to 95 per cent of the gross or net level of emissions and those that are within the top 95 per cent of the categories that contribute to the change between 2000 and 2012, or the trend of emissions. the keycategoriesidentifiedin2012aresummarisedinTable 1.8. in accordance with the 2006 ipCC Guidelines, the key category analysis is performed once for the inventory excluding the FolU sector and then repeated for the inventory including the FolU sector.

the level analysis excluding FolU (Table 1.9)identifiesthe following major contributions to the net emissions for 2012:

• Co2 emissions from solid fuels in energy industries – main activity electricity and heat (47.9%);

• Co2 emissions from liquid fuels used in road transportation (8.5%);

• Co2 emissions from liquid fuels in energy industries – manufacture of solid fuels and other energy industries (5.9%);

• Co2 emissions from the use of solid fuels in manufacturing industries and construction (4.8%); and

• CH4 emissions from enteric fermentation from cattle (4.2%).

including FolU in the level assessment (Table 1.10) shows the following key categories:

• Co2 emissions from solid fuels in energy industries – main activity electricity and heat (43.5%);

• Co2 emissions from liquid fuels used in road transportation (7.7%);

• Co2 emissions from liquid fuels in energy industries – manufacture of solid fuels and other energy industries (5.4%);

• Co2 removals from land converted to forest land (5.3%); and

• Co2 emissions from the use of solid fuels in manufacturing industries and construction (4.4%).

Table 1.11identifiesthekeycategories(excludingFOLU)havingthelargestrelativeinfluenceonthetrend,whencompared with the average change in net emissions from 2000 to 2012:

• Co2 emissions from solid fuels used in other sectors – residential (14.7%);

• Co2 emissions from solid fuels used in energy industries – main activity electricity and heat production (13.0%);

• Co2 emissions from solid fuels used in manufacturing industries and construction (12.3%);

• Co2 emissions from liquid fuels used in energy industries – manufacture of solid fuels and other energy industries (8.3%); and

1. introduction


table 1.8: Summary of South africa’s key categories for 2012 level assessment and the trend assessment for 2000 to 2012 (including and excluding FolU activities).

IPCC Category code

IPCC CategoryGreenhouse

gasCriteria for

identification

ENERGY

1.a.1.aenergy industries - Solid Fuels - Main activity electricity and heat production

Co2 l1, t1

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining Co2 l1, t1

1.a.1.cenergy industries - liquid Fuels - Manufacture of solid fuels and other energy industries

Co2 l1, t1

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 l1, t1

1.a.2Manufacturing industries and Construction - Gaseous Fuels

Co2 l1

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 l1, t1

1.a.2 Manufacturing industries and Construction - liquid Fuels Co2 l1, t1

1.a.3.a transport - liquid Fuels - Civil aviation Co2 l1, t1

1.a.3.b transport - liquid Fuels - road transportation Co2 l1, t1

1.a.3.c transport - liquid Fuels - railways Co2 t1

1.a.4.a other Sectors - liquid Fuels - Commercial/institutional Co2 l1, t1

1.a.4.a other Sectors - Solid Fuels - Commercial/institutional Co2 l1, t1

1.a.4.b other Sectors - Solid Fuels - residential Co2 l1, t1

1.a.4.b other Sectors - liquid Fuels - residential Co2 l1, t1

1.a.4.cother Sectors - liquid Fuels - agriculture/Forestry/Fish-ing/Fish farms

Co2 l1, t1

• CH4 emissions from enteric fermentation from cattle (7.5%).

including FolU in the trend analysis (table 1.12) also shows the following key categories:

• Co2 removals from land converted to forest land (14.6%);

• Co2 emissions from solid fuels used in energy industries – main activity electricity and heat production (14.1%);

• Co2 emissions from solid fuels used in other sectors – residential (11.7%);

• Co2 emissions from solid fuels used in manufacturing industries and construction (9.1%); and

• Co2 emissions from liquid fuels used in energy industries – manufacture of solid fuels and other energy industries (6.0%).

the complete level and trend assessments are provided in Appendix b.


IPCC Category code

IPCC CategoryGreenhouse

gasCriteria for

identification

IPPU

2.a.1 Mineral industry - Cement production Co2 l1

2.B. Chemical industry - industry x Co2 t1

2.B. Chemical industry - industry F N2o t1

2.B. Chemical industry - industry K Co2 t1

2.C.1 Metal industry - iron and Steel production Co2 l1, t1

2.C.2 Metal industry - Ferroalloys production Co2 l1, t1

2.C.3 Metal industry - aluminium production pFCs t1

2.F.1product uses as substitutes for oDS - refrigeration and air Conditioning

HFCs t1

AFOLU

3.a.1.a enteric Fermentation - Cattle CH4 l1, t1

3.a.1.c enteric Fermentation - Sheep CH4 l1, t1

3.a.1.d enteric Fermentation - Goats CH4 t1

3.B.1.a Forest land - Forest land remaining Forest land Co2 l1

3.B.1.b Forest land - land Converted to Forest land Co2 l1, t1

3.B.2.a Cropland - Cropland remaining Cropland Co2 t1

3.B.2.b Cropland - land Converted to Cropland Co2 l1, t1

3.B.3.b Grassland - land Converted to Grassland Co2 l1, t1

3.C.4Direct N2o emissions from managed soils - Urine and dung N deposits in prp

N2o l1, t1

3.C.4Direct N2o emissions from managed soils - inorganic N application

N2o t1

3.C.5indirect N2o emissions from managed soils - N leaching/runoff

N2o l1, t1

3.C.5indirect N2o emissions from managed soils - N volatilization

N2o t1

WASTE

4.a Solid Waste Disposal CH4 l1, t1

4.D.1 Wastewater treatment and Discharge - Domestic CH4 l1

1. introduction


table 1.9: level assessment results for 2012, excluding FolU contributions. only key categories are shown.

IPCC code

IPCC Category GHG2012

estimate(Gg CO2 eq)

Level assessment

(%)

Cumulative total (%)

1.a.1,aenergy industries - Solid Fuels - Main activity electricity and heat production

Co2 242 154.0 47.9 47.9

1.a.3.btransport - liquid Fuels - road transportation

Co2 42 929.6 8.5 56.4

1.a.1.cenergy industries - liquid Fuels - Man-ufacture of solid fuels and other energy industries

Co2 29 903.3 5.9 62.3

1.a.2Manufacturing industries and Construc-tion - Solid Fuels

Co2 24 472.7 4.8 67.2

3.a.1.a enteric Fermentation – Cattle CH4 21 236.5 4.2 71.4

1.a.4.b other Sectors - Solid Fuels – residential Co2 18 461.9 3.7 75.0

4.a Solid Waste Disposal CH4 18 412.0 3.6 78.7

2.C.1 Metal industry - iron and Steel production Co2 15 020.7 3.0 81.6

1.a.4.aother Sectors - liquid Fuels - Commer-cial/institutional

Co2 12 685.4 2.5 84.1


N2o 11 815.0 2.3 86.5

2.C.2 Metal industry - Ferroalloys production Co2 11 623.8 2.3 88.8

2.a.1 Mineral industry - Cement production Co2 3 844.5 0.8 89.5

3.a.1.c enteric Fermentation – Sheep CH4 3 783.5 0.7 90.3

1.a.3.a transport - liquid Fuels - Civil aviation Co2 3 466.4 0.7 91.0

1.a.4.aother Sectors - Solid Fuels - Commercial/institutional

Co2 3 434.8 0.7 91.7

1.a.4.cother Sectors - liquid Fuels - agriculture/Forestry/Fishing/Fish farms

Co2 3 420.3 0.7 92.3

4.D.1Wastewater treatment and Discharge - Domestic

CH4 2 827.7 0.6 92.9


IPCC code


estimate(Gg CO2 eq)

Level assessment

(%)


1.a.1.benergy industries - liquid Fuels - Petroleumrefining

Co2 2 802.4 0.6 93.4


Co2 2 497.5 0.5 93.9


N2o 2 330.0 0.5 94.4

1.a.4.b other Sectors - liquid Fuels - residential Co2 2 154.6 0.4 94.8

1.a.2Manufacturing industries and Construction - liquid Fuels

Co2 2 114.0 0.4 95.2

table 1.10: level assessment results for 2012, including FolU contributions. only key categories are shown.

IPCC code


estimate(Gg CO2 eq)

Level assessment

(%)



Co2 242 154.0 43.5 43.5


Co2 42 929.6 7.7 51.2


Co2 29 903.3 5.4 56.6

3.B.1.bForest land - land Converted to Forest land

Co2 -29 490.8 5.3 61.9

1.a.2Manufacturing industries and Construction - Solid Fuels

Co2 24 472.7 4.4 66.3

3.a.1.a enteric Fermentation - Cattle CH4 21 236.5 3.8 70.1

1.a.4.b other Sectors - Solid Fuels - residential Co2 18 461.9 3.3 73.4



1. introduction


IPCC code


estimate(Gg CO2 eq)

Level assessment

(%)


1.a.4.aother Sectors - liquid Fuels - Commer-cial/institutional

Co2 12 685.4 2.3 81.7


N2o 11 815.0 2.1 83.9


3.B.3.b Grassland - land Converted to Grassland Co2 7 871.8 1.4 87.4

3.B.2.b Cropland - land Converted to Cropland Co2 5 721.0 1.0 88.4

3.B.1.aForest land - Forest land remaining Forest land

Co2 -4 834.4 0.9 89.3


3.a.1.c enteric Fermentation - Sheep CH4 3 783.5 0.7 90.6



Co2 3 434.8 0.6 91.9


Co2 3 420.3 0.6 92.5


CH4 2 827.7 0.5 93.0


Co2 2 802.4 0.5 93.5


Co2 2 497.5 0.4 93.9


N2o 2 330.0 0.4 94.4



Co2 2 114.0 0.4 95.1


tabl

e 1.

11:

tren

d as

sess

men

t re

sult

s fo

r 20

12 (w

ith

2000

as

the

base

yea

r), e

xclu

ding

Fo

lU c

ontr

ibut

ions

. onl

y th

e ke

y ca

tego

ries

are

sho

wn.

IPC

C

code

IPC

C C

ateg

ory

GH

G20

00 Y

ear

esti

mat

e(G

g C

o2e

q)

2012

Yea

r E

stim

ate

(Gg

Co

2eq)

Tren

d A

sses

smen

t

Con

tri-

buti

on t

o tr

end

(%)

Cum

ula-

tive

tot

al

(%)

1.a

.4.b

oth

er S

ecto

rs -

Sol

id F

uels

-

res

iden

tial

Co

23

604.

218

461

.90.

034

14.7

14.7

1.a

.1,a

ener

gy in

dust

ries

- S

olid

Fue

ls -

Mai

n ac

tivity

ele

ctri

city

and

hea

t pr

oduc

tion

Co

218

5 02

7.4

242

154.

00.

030

13.0

27.7

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

Solid

Fue

lsC

o2

29 1

01.6

24 4

72.7

0.02

912

.340

.0

1.a

.1.c

ener

gy in

dust

ries

- l

iqui

d Fu

els

- M

anuf

actu

re o

f sol

id fu

els

and

othe

r en

ergy

indu

stri

esC

o2

30 4

54.7

29 9

03.3

0.01

98.

348

.3

3.a

.1.a

ente

ric

Ferm

enta

tion

– C

attle

CH

422

819

.421

236

.50.

017

7.5

55.8

4.a

Solid

Was

te D

ispo

sal

CH

49

367.

618

412

.00.

017

7.1

62.9

2.C

.1M

etal

indu

stry

- ir

on a

nd S

teel

pro

-du

ctio

nC

o2

16 4

10.5

15 0

20.7

0.01

35.

668

.5

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

Uri

ne a

nd d

ung

N d

epos

its in

pr

pN

2o13

180

.811

815

.00.

011

4.8

73.3

1.a

.4.a

oth

er S

ecto

rs -

liq

uid

Fuel

s -

Com

mer

cial

/inst

itutio

nal

Co

27

690.

412

685

.40.

008

3.3

76.6

1.a

.3.b

tran

spor

t -

liqu

id F

uels

- r

oad

tran

spor

tatio

nC

o2

32 6

23.3

42 9

29.6

0.00

62.

579

.2

2.C

.2M

etal

indu

stry

- F

erro

allo

ys

prod

uctio

nC

o2

8 07

9.1

11 6

23.8

0.00

41.

780

.8

1.a

.4.b

oth

er S

ecto

rs -

liq

uid

Fuel

s –

res

iden

tial

Co

22

868.

92

154.

60.

003

1.5

82.3

2.F.1

prod

uct

uses

as

subs

titut

es fo

r o

DS

- r

efri

gera

tion

and

air

Con

ditio

ning

HFC

s0.

01

396.

10.

003

1.5

83.8

3.a

.1.c

ente

ric

Ferm

enta

tion

- Sh

eep

CH

44

169.

03

783.

50.

003

1.5

85.3

1.a

.4.a

oth

er S

ecto

rs -

Sol

id F

uels

-

Com

mer

cial

/inst

itutio

nal

Co

21

811.

23

434.

80.

003

1.2

86.5

1.a

.1.b

ener

gy in

dust

ries

- l

iqui

d Fu

els

- Petroleumrefining

Co

23

164.

82

802.

40.

003

1.2

87.7

1.a

.3.a

tran

spor

t -

liqu

id F

uels

- C

ivil

avi

a-tio

nC

o2

2 04

0.0

3 46

6.4

0.00

21.

088

.7

2.C

.3M

etal

indu

stry

- a

lum

iniu

m p

rodu

ctio

npF

Cs

982.

21

979.

20.

002

0.8

89.5

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

leac

hing

/run

off

N2o

2 47

9.0

2 33

0.0

0.00

20.

890

.3

2.B

Che

mic

al in

dust

ry -

indu

stry

FN

2oN

Sa N

Sa0.

002

0.8

91.0

1.a

.4.c

oth

er S

ecto

rs -

liq

uid

Fuel

s -

agr

icul

ture

/For

estr

y/Fi

shin

g/Fi

sh fa

rms

Co

22

207.

23

420.

30.

002

0.7

91.8

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

liqu

id F

uels

Co

21

186.

12

114.

00.

002

0.7

92.4

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

vol

atili

zatio

nN

2o1

939.

51

791.

70.

002

0.6

93.1

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

inor

gani

c N

app

licat

ion

N2o

1 93

4.7

2 00

0.1

0.00

10.

493

.5

2.B

Che

mic

al in

dust

ry -

indu

stry

xC

o2

NSa

NSa

0.00

10.

493

.9

2.B

Che

mic

al in

dust

ry -

indu

stry

KC

o2

NSa

NSa

0.00

10.

494

.3

3.a

.1.d

ente

ric

Ferm

enta

tion

- G

oats

CH

499

3.6

855.

60.

001

0.4

94.7

1.a

.3.c

tran

spor

t -

liqu

id F

uels

- r

ailw

ays

Co

255

1.5

318.

00.

001

0.4

95.1

1. introduction


IPC

C

code

IPC

C C

ateg

ory

GH

G20

00 Y

ear

esti

mat

e(G

g C

o2e

q)

2012

Yea

r E

stim

ate

(Gg

Co

2eq)

Tren

d A

sses

smen

t

Con

tri-

buti

on t

o tr

end

(%)

Cum

ula-

tive

tot

al

(%)

1.a

.4.b

oth

er S

ecto

rs -

Sol

id F

uels

-

res

iden

tial

Co

23

604.

218

461

.90.

034

14.7

14.7

1.a

.1,a

ener

gy in

dust

ries

- S

olid

Fue

ls -

Mai

n ac

tivity

ele

ctri

city

and

hea

t pr

oduc

tion

Co

218

5 02

7.4

242

154.

00.

030

13.0

27.7

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

Solid

Fue

lsC

o2

29 1

01.6

24 4

72.7

0.02

912

.340

.0

1.a

.1.c

ener

gy in

dust

ries

- l

iqui

d Fu

els

- M

anuf

actu

re o

f sol

id fu

els

and

othe

r en

ergy

indu

stri

esC

o2

30 4

54.7

29 9

03.3

0.01

98.

348

.3

3.a

.1.a

ente

ric

Ferm

enta

tion

– C

attle

CH

422

819

.421

236

.50.

017

7.5

55.8

4.a

Solid

Was

te D

ispo

sal

CH

49

367.

618

412

.00.

017

7.1

62.9

2.C

.1M

etal

indu

stry

- ir

on a

nd S

teel

pro

-du

ctio

nC

o2

16 4

10.5

15 0

20.7

0.01

35.

668

.5

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

Uri

ne a

nd d

ung

N d

epos

its in

pr

pN

2o13

180

.811

815

.00.

011

4.8

73.3

1.a

.4.a

oth

er S

ecto

rs -

liq

uid

Fuel

s -

Com

mer

cial

/inst

itutio

nal

Co

27

690.

412

685

.40.

008

3.3

76.6

1.a

.3.b

tran

spor

t -

liqu

id F

uels

- r

oad

tran

spor

tatio

nC

o2

32 6

23.3

42 9

29.6

0.00

62.

579

.2

2.C

.2M

etal

indu

stry

- F

erro

allo

ys

prod

uctio

nC

o2

8 07

9.1

11 6

23.8

0.00

41.

780

.8

1.a

.4.b

oth

er S

ecto

rs -

liq

uid

Fuel

s –

res

iden

tial

Co

22

868.

92

154.

60.

003

1.5

82.3

2.F.1

prod

uct

uses

as

subs

titut

es fo

r o

DS

- r

efri

gera

tion

and

air

Con

ditio

ning

HFC

s0.

01

396.

10.

003

1.5

83.8

3.a

.1.c

ente

ric

Ferm

enta

tion

- Sh

eep

CH

44

169.

03

783.

50.

003

1.5

85.3

1.a

.4.a

oth

er S

ecto

rs -

Sol

id F

uels

-

Com

mer

cial

/inst

itutio

nal

Co

21

811.

23

434.

80.

003

1.2

86.5

1.a

.1.b

ener

gy in

dust

ries

- l

iqui

d Fu

els

- Petroleumrefining

Co

23

164.

82

802.

40.

003

1.2

87.7

1.a

.3.a

tran

spor

t -

liqu

id F

uels

- C

ivil

avi

a-tio

nC

o2

2 04

0.0

3 46

6.4

0.00

21.

088

.7

2.C

.3M

etal

indu

stry

- a

lum

iniu

m p

rodu

ctio

npF

Cs

982.

21

979.

20.

002

0.8

89.5

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

leac

hing

/run

off

N2o

2 47

9.0

2 33

0.0

0.00

20.

890

.3

2.B

Che

mic

al in

dust

ry -

indu

stry

FN

2oN

Sa N

Sa0.

002

0.8

91.0

1.a

.4.c

oth

er S

ecto

rs -

liq

uid

Fuel

s -

agr

icul

ture

/For

estr

y/Fi

shin

g/Fi

sh fa

rms

Co

22

207.

23

420.

30.

002

0.7

91.8

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

liqu

id F

uels

Co

21

186.

12

114.

00.

002

0.7

92.4

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

vol

atili

zatio

nN

2o1

939.

51

791.

70.

002

0.6

93.1

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

inor

gani

c N

app

licat

ion

N2o

1 93

4.7

2 00

0.1

0.00

10.

493

.5

2.B

Che

mic

al in

dust

ry -

indu

stry

xC

o2

NSa

NSa

0.00

10.

493

.9

2.B

Che

mic

al in

dust

ry -

indu

stry

KC

o2

NSa

NSa

0.00

10.

494

.3

3.a

.1.d

ente

ric

Ferm

enta

tion

- G

oats

CH

499

3.6

855.

60.

001

0.4

94.7

1.a

.3.c

tran

spor

t -

liqu

id F

uels

- r

ailw

ays

Co

255

1.5

318.

00.

001

0.4

95.1

a N

ot s

how

n as

the

dat

a is

sen

sitiv

e an

d th

e di

sagg

rega

ted

chem

ical

indu

stry

[2B

] da

ta is

not

rep

orte

d.


tabl

e 1.

12:

tren

d as

sess

men

t re

sult

s fo

r 20

12 (w

ith

2000

as

the

base

yea

r), i

nclu

ding

Fo

lU c

ontr

ibut

ions

. o

nly

the

key

cate

gori

es a

re s

how

n.

IPC

C

code

IPC

C C

ateg

ory

GH

G20

00 Y

ear

esti

mat

e(G

g C

o2e

q)

2012

Yea

r E

stim

ate

(Gg

Co

2eq)

Tren

d A

sses

smen

t

Con

tri-

buti

on t

o tr

end

(%)

Cum

ula-

tive

tot

al

(%)

3.B.

1.b

Fore

st la

nd -

lan

d C

onve

rted

to

Fore

st la

ndC

o2

-153

39.9

-294

90.8

0.03

914

.614

.6

1.a

.1,a

ener

gy in

dust

ries

- S

olid

Fue

ls -

Mai

n ac

tivity

ele

ctri

city

and

hea

t pr

oduc

tion

Co

218

5027

.424

2154

.00.

038

14.1

28.6

1.a

.4.b

oth

er S

ecto

rs -

Sol

id F

uels

-

res

iden

tial

Co

236

04.2

1846

1.9

0.03

211

.740

.3

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

Solid

Fue

lsC

o2

2910

1.6

2447

2.7

0.02

59.

149

.5

1.a

.1.c

ener

gy in

dust

ries

- l

iqui

d Fu

els

- M

anuf

actu

re o

f sol

id fu

els

and

othe

r en

ergy

indu

stri

esC

o2

3045

4.7

2990

3.3

0.01

66.

055

.4

4.a

Solid

Was

te D

ispo

sal

CH

493

67.6

1841

2.0

0.01

65.

861

.3

3.a

.1.a

ente

ric

Ferm

enta

tion

- C

attle

CH

422

819.

421

236.

50.

015

5.4

66.7

2.C

.1M

etal

indu

stry

- ir

on a

nd S

teel

pr

oduc

tion

Co

216

410.

515

020.

70.

011

4.1

70.8

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

Uri

ne a

nd d

ung

N d

epos

its in

pr

pN

2o13

180.

811

815.

00.

009

3.5

74.3

1.a

.4.a

oth

er S

ecto

rs -

liq

uid

Fuel

s -

Com

mer

cial

/inst

itutio

nal

Co

276

90.4

1268

5.4

0.00

72.

877

.1

1.a

.3.b

tran

spor

t -

liqu

id F

uels

- r

oad

tran

spor

tatio

nC

o2

3262

3.3

4292

9.6

0.00

72.

779

.8

2.C

.2M

etal

indu

stry

- F

erro

allo

ys

prod

uctio

nC

o2

8079

.111

623.

80.

004

1.5

81.3

3.B.

3.b

Gra

ssla

nd -

lan

d C

onve

rted

to

Gra

ssla

ndC

o2

7904

.178

71.8

0.00

41.

582

.7

2.F.1

prod

uct

uses

as

subs

titut

es fo

r o

DS

- r

efri

gera

tion

and

air

Con

ditio

ning

HFC

s0.

013

96.1

0.00

31.

283

.9

1.a

.4.b

oth

er S

ecto

rs -

liq

uid

Fuel

s –

res

iden

tial

Co

228

68.9

2154

.60.

003

1.1

85.0

3.a

.1.c

ente

ric

Ferm

enta

tion

- Sh

eep

CH

441

69.0

3783

.50.

003

1.1

86.1

3.B.

2.b

Cro

plan

d -

land

Con

vert

ed t

o C

ropl

and

Co

257

21.7

5721

.00.

003

1.0

87.1

1.a

.4.a

oth

er S

ecto

rs -

Sol

id F

uels

-

Com

mer

cial

/inst

itutio

nal

Co

218

11.2

3434

.80.

003

1.0

88.1

1.a

.1.b

ener

gy in

dust

ries

- l

iqui

d Fu

els

- Petroleumrefining

Co

231

64.8

2802

.40.

002

0.9

89.0

1.a

.3.a

tran

spor

t -

liqu

id F

uels

- C

ivil

avi

atio

nC

o2

2040

.034

66.4

0.00

20.

889

.8

3.B.

2.a

Cro

plan

d -

Cro

plan

d r

emai

ning

C

ropl

and

Co

2-1

241.

6-1

04.9

0.00

20.

790

.6

2.C

.3M

etal

indu

stry

- a

lum

iniu

m p

rodu

ctio

npF

Cs

982.

219

79.2

0.00

20.

791

.2

2.B

Che

mic

al in

dust

ry -

indu

stry

FN

2oN

Sa N

Sa0.

002

0.6

91.8

1.a

.4.c

oth

er S

ecto

rs -

liq

uid

Fuel

s -

agr

icul

ture

/For

estr

y/Fi

shin

g/Fi

sh fa

rms

Co

222

07.2

3420

.30.

002

0.6

92.4

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

leac

hing

/run

off

N2o

2479

.023

30.0

0.00

20.

693

.0

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

liqu

id F

uels

Co

211

86.1

2114

.00.

002

0.6

93.6

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

vol

atili

zatio

nN

2o19

39.5

1791

.70.

001

0.5

94.0

2.B.

6C

hem

ical

indu

stry

- in

dust

ry K

Co

2N

Sa N

Sa0.

001

0.3

94.4

2.B.

1C

hem

ical

indu

stry

- in

dust

ry x

Co

2N

Sa N

Sa 0.

001

0.3

94.7

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

inor

gani

c N

app

licat

ion

N2 o

1934

.720

00.1

0.00

10.

395

.0

3.a

.1.d

ente

ric

Ferm

enta

tion

- G

oats

CH

499

3.6

855.

60.

001

0.3

95.3

1. introduction


IPC

C

code

IPC

C C

ateg

ory

GH

G20

00 Y

ear

esti

mat

e(G

g C

o2e

q)

2012

Yea

r E

stim

ate

(Gg

Co

2eq)

Tren

d A

sses

smen

t

Con

tri-

buti

on t

o tr

end

(%)

Cum

ula-

tive

tot

al

(%)

3.B.

1.b

Fore

st la

nd -

lan

d C

onve

rted

to

Fore

st la

ndC

o2

-153

39.9

-294

90.8

0.03

914

.614

.6

1.a

.1,a

ener

gy in

dust

ries

- S

olid

Fue

ls -

Mai

n ac

tivity

ele

ctri

city

and

hea

t pr

oduc

tion

Co

218

5027

.424

2154

.00.

038

14.1

28.6

1.a

.4.b

oth

er S

ecto

rs -

Sol

id F

uels

-

res

iden

tial

Co

236

04.2

1846

1.9

0.03

211

.740

.3

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

Solid

Fue

lsC

o2

2910

1.6

2447

2.7

0.02

59.

149

.5

1.a

.1.c

ener

gy in

dust

ries

- l

iqui

d Fu

els

- M

anuf

actu

re o

f sol

id fu

els

and

othe

r en

ergy

indu

stri

esC

o2

3045

4.7

2990

3.3

0.01

66.

055

.4

4.a

Solid

Was

te D

ispo

sal

CH

493

67.6

1841

2.0

0.01

65.

861

.3

3.a

.1.a

ente

ric

Ferm

enta

tion

- C

attle

CH

422

819.

421

236.

50.

015

5.4

66.7

2.C

.1M

etal

indu

stry

- ir

on a

nd S

teel

pr

oduc

tion

Co

216

410.

515

020.

70.

011

4.1

70.8

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

Uri

ne a

nd d

ung

N d

epos

its in

pr

pN

2o13

180.

811

815.

00.

009

3.5

74.3

1.a

.4.a

oth

er S

ecto

rs -

liq

uid

Fuel

s -

Com

mer

cial

/inst

itutio

nal

Co

276

90.4

1268

5.4

0.00

72.

877

.1

1.a

.3.b

tran

spor

t -

liqu

id F

uels

- r

oad

tran

spor

tatio

nC

o2

3262

3.3

4292

9.6

0.00

72.

779

.8

2.C

.2M

etal

indu

stry

- F

erro

allo

ys

prod

uctio

nC

o2

8079

.111

623.

80.

004

1.5

81.3

3.B.

3.b

Gra

ssla

nd -

lan

d C

onve

rted

to

Gra

ssla

ndC

o2

7904

.178

71.8

0.00

41.

582

.7

2.F.1

prod

uct

uses

as

subs

titut

es fo

r o

DS

- r

efri

gera

tion

and

air

Con

ditio

ning

HFC

s0.

013

96.1

0.00

31.

283

.9

1.a

.4.b

oth

er S

ecto

rs -

liq

uid

Fuel

s –

res

iden

tial

Co

228

68.9

2154

.60.

003

1.1

85.0

3.a

.1.c

ente

ric

Ferm

enta

tion

- Sh

eep

CH

441

69.0

3783

.50.

003

1.1

86.1

3.B.

2.b

Cro

plan

d -

land

Con

vert

ed t

o C

ropl

and

Co

257

21.7

5721

.00.

003

1.0

87.1

1.a

.4.a

oth

er S

ecto

rs -

Sol

id F

uels

-

Com

mer

cial

/inst

itutio

nal

Co

218

11.2

3434

.80.

003

1.0

88.1

1.a

.1.b

ener

gy in

dust

ries

- l

iqui

d Fu

els

- Petroleumrefining

Co

231

64.8

2802

.40.

002

0.9

89.0

1.a

.3.a

tran

spor

t -

liqu

id F

uels

- C

ivil

avi

atio

nC

o2

2040

.034

66.4

0.00

20.

889

.8

3.B.

2.a

Cro

plan

d -

Cro

plan

d r

emai

ning

C

ropl

and

Co

2-1

241.

6-1

04.9

0.00

20.

790

.6

2.C

.3M

etal

indu

stry

- a

lum

iniu

m p

rodu

ctio

npF

Cs

982.

219

79.2

0.00

20.

791

.2

2.B

Che

mic

al in

dust

ry -

indu

stry

FN

2oN

Sa N

Sa0.

002

0.6

91.8

1.a

.4.c

oth

er S

ecto

rs -

liq

uid

Fuel

s -

agr

icul

ture

/For

estr

y/Fi

shin

g/Fi

sh fa

rms

Co

222

07.2

3420

.30.

002

0.6

92.4

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

leac

hing

/run

off

N2o

2479

.023

30.0

0.00

20.

693

.0

1.a

.2M

anuf

actu

ring

indu

stri

es a

nd

Con

stru

ctio

n -

liqu

id F

uels

Co

211

86.1

2114

.00.

002

0.6

93.6

3.C

.5in

dire

ct N

2o e

mis

sion

s fr

om m

anag

ed

soils

- N

vol

atili

zatio

nN

2o19

39.5

1791

.70.

001

0.5

94.0

2.B.

6C

hem

ical

indu

stry

- in

dust

ry K

Co

2N

Sa N

Sa0.

001

0.3

94.4

2.B.

1C

hem

ical

indu

stry

- in

dust

ry x

Co

2N

Sa N

Sa 0.

001

0.3

94.7

3.C

.4D

irect

N2o

em

issi

ons

from

man

aged

so

ils -

inor

gani

c N

app

licat

ion

N2 o

1934

.720

00.1

0.00

10.

395

.0

3.a

.1.d

ente

ric

Ferm

enta

tion

- G

oats

CH

499

3.6

855.

60.

001

0.3

95.3

a N

ot s

how

n as

the

dat

a is

sen

sitiv

e an

d th

e di

sagg

rega

ted

chem

ical

indu

stry

[2B

] da

ta is

not

rep

orte

d.


1.4 Information on QA/QC plan

in accordance with ipCC requirements for national GHG inventory preparation, the necessary quality control and quality assurance (QC/Qa) measures for emissions reporting should be summarised in a QC/Qa plan. the primary purpose of a QC/Qa plan is to organise, plan and monitor QC/Qa measures. the objective of quality checking is to improve the transparency, consistency, comparability, completeness, and accuracy of the national GHG inventory. the basic requirements of QC/Qa

assurance measures for national GHG inventories are definedinthe2006IPCCGuidelines,Vol.1,Chapter6.

1.4.1 Quality control

the quality controls applied in this GHG inventory involved generic quality checks related to the calculations, data processing, and completeness and documents applicable to the inventory. the QC procedures applied in this inventory are summarised in Table 1.13.

table 1.13: QC activity and procedures applied in this inventory

QC Activity Procedures

activity data QC• Check the temporal consistency of the activity data; and • Check the consistency of the units.

eF data QC

• ipCC default eF: o Check default eF applicability; o Check temporal consistency; and o Check the consistency of the units.

• CountryspecificEF: o Check data sources for CS eF; o Check whether QC procedures were conducted on the CS eF; o Check QC of models; o Compare CS eF to ipCC default eF; and o Compare CS eF to other country eF

General data QC

• Check the data calculations: o reproduce a set of emission/removal calculations; and o Check calculations with an approximation method.

• Check any recalculation data;• Check the emission and removal data are correctly aggregated from

lower reporting levels;• Check that the data is compared to previous estimates;• Check for consistency in the trend; and• Check for completeness of each subcategory

Uncertainty QC• Check that expert judgement is recorded; and• Check uncertainty calculations

1. introduction


1.4.2 Quality assurance (QA)

the Qa process was done by external reviewers who have a different technical perspective and who were not involved in the compilation process of the inventory, so they could conduct an unbiased review of the inventory. the external reviewers ensured that the inventory’s results, assumptions and methods were reasonable. Furthermore, a public review process was undertaken to supplement the external review. the independent review process is demonstrated in Figure 1.3.

essentially, the draft GHG inventory report was published for public comment in a Government Gazette. parallel to the publication for public comment, an independent technical review of activity data, emission factors and methodologies used was undertaken by an independent company. Comments that were submitted during the public commenting process were incorporated in the independent review process. Findings and recommendations from the independent review process were then used to refine the draft GHG inventory report. Currently, the Dea is preparing a revised

QC Activity Procedures

Database QC

• Check that the data is in the database;• Check for transcription errors;• Check uncertainty is in the database;• Check for transcription errors in uncertainty data;• Check the correct units have been used in the database;• Check the labels in the database are correct;• Check for common data consistency between sectors and categories;• Check the correct conversion factors are used;• Check data aggregations are correct; and• Check the uncertainty aggregations are correct.

Metadata QC

• Check all data details are documented;• Check all uncertainty details are documented;• Check all recalculation details are recorded;• Check all supporting data and documents are attached; and• Check details regarding experts are documented.

report QC

• Check that the relevant activity data details are incorporated into the report;

• Check that the relevant eF data details are incorporated into the report;• Check for data transcription errors between the database and the report;• Check the references are properly cited;• Check that expert judgements are reported;• Check that uncertainty data in incorporated into the report;• Check that completeness is documented;• Check that comparisons have been made with previous estimates;• Check that explanations are provided for any data differences; and• Check that the QC/Qa procedures have been documented.


three-year supplementary improvement plan to address recommendations from the independent review process. Some of the projects that are currently listed in the Greenhouse Gas improvement programme (GHGip) areasaresultofthefindingsofthepreviousreviewundertaken for the 2000 – 2010 GHG inventory.

in addition to the ipCC Guidelines, South africa has developeditsownvalidationandverificationprocedurefor GHG assertions for corporate reporting of emissions, and for emission estimation linked to voluntary market schemes aimed at reducing emissions. a GHG assertion is definedasadeclarationorfactualandobjectivestatementmade by a person or persons responsible for the GHG inventory and the supporting GHG information. it is a standard adopted from the international Standardization

organization (iSo) series programme for data documentation and audits as part of a quality management system.

in the South african context, Qa/QC measures are definedbyPart3oftheSouthAfricanNationalStandardforGreenhouseGases,SANS14064-3:2006(SpecificationwithguidanceforthevalidationandverificationofGHGgasassertions).Thisstandardspecifiestherequirementsfor selecting GHG validators/verifiers, establishing the level of assurance, objectives, criteria and scope, determiningthevalidation/verificationapproach,assessingGHG data, information, information systems and controls, evaluatingassertions,andpreparingvalidation/verificationstatements.

Figure 1.3: the independent review process for the 2000 – 2012 inventory.

Draft GHG inventory

Comments from the public

Comment and Response Report

Pub

lic C

omm

ent

Pro

cess

Inde

pend

ent

Rev

iew GHG Inventory

Review Report

3-year Supplementary Improvement

Plan

GHG Inventory Summary

Report

Comments and response

database

1. introduction


1.5 Evaluating uncertainty

Uncertainty analysis is regarded by the ipCC Guidelines as an essential element of any complete GHG inventory. Uncertainty reporting is important for suggesting methods of compiling uncertainty estimates and for identifying approaches that would enable the prioritisation of national efforts to reduce future uncertainties. it identifiesareasoffurtherimprovementintheinventorypreparation process and will guide methodological choices in future inventories to improve accuracy, reduce bias, and transparently report on the presence and levels of uncertainty. Uncertainty can also be used to determine appropriate methods for carrying out recalculations for previous inventories. the overall objective is to minimise uncertainties to the greatest possible degree.

Chapter 3 of the 2006 ipCC Guidelines for National Greenhouse Gas inventories describes the need and methodology for estimating and reporting uncertainties associated with annual estimates of emissions and removals, including emission and removal trends over time. Broadly speaking, the approach involves combining the category uncertainties into estimates of uncertainty for total national net emissions and their associated trends.

Uncertainty analysis for the South africa inventory involved quantifying the uncertainties for all source categories and sinks where data were available. the main sources for the uncertainty in activity and emission factor data are the ipCC default uncertainty estimates (particularly for theemissionfactors),scientificliteratureandreportedvariances, expert judgement and comparisons with uncertainty ranges reported by other countries.

there is a lack of quantitative uncertainty data in many sourcecategories,whichmakesitdifficulttodetermineand assess the uncertainty of these emission sources. in many cases more detailed data collection is required to speculate about levels of quantitative uncertainty. Wheresufficientdatawereavailable,theanalysisinvolvedthe determination of a probability-density function for

a number of parameters, using approaches and values provided in the 2006 ipCC Guidelines. thus, the uncertainty analysis included a statistical evaluation of individual data items, and experts’ assessments as guided by the ipCC Good practice Guidelines.

For this inventory an approach 1 uncertainty analysis (assuming no correlation between the activity data) was applied to the energy and the ippU sectors and the ipCC uncertaintytableswereused.Therewereinsufficientdata to include aFolU and waste sector uncertainties, however these will be included in the next inventory. the energy sector was determined to have an overall uncertainty of 6.5%, while the uncertainty introduced into the trend was 6.3%. the ippU sector, on the other hand, had an uncertainty of 30.6%. these uncertainties are elevated due to the incorporation of sub-section 2F (product uses as substitutes for ozone-depleting substances) which has no emission estimates for the years 2000 to 2004 and then quite high emissions estimated in the years between 2005 and 2012. if this category is excluded from the uncertainty analysis then the total uncertainty drops to 8.8%, and the trend uncertainty is reduced to 4.74%. the detailed analyses are provided in Appendix C.

1.6 General assessment of completeness

the South african GHG emission inventory for the period 2000 – 2012 is not complete, mainly due to the lack of sufficientdata.Table 1.14identifiesthesourcesinthe2006 ipCC Guidelines which were not included in this inventory and the reason for their omissions is discussed further in the appropriate chapters.


table 1.14: activities in the 2012 inventory which are not estimated (Ne), included elsewhere (ie) or not occurring (No).

NE, IE or NO

Activity Comments

NE

Co2 and CH4 fugitive emissions from oil and natural gas operations

emissions from this source category will be included in the next inventory submission covering the period 2000-2014

Co2, CH4 and N2o from spontaneous combustion of coal seams

New research work on sources of emissions from this category will be used to report emissions in the next inventory submission

CH4 emissions from abandoned minesNew research work on sources of emissions from this category will be used to report emissions in the next inventory submission

other process use of carbonates

electronics industrya study was to be undertaken in 2015 to understand emissions from this source category

indirect N2o emission due to nitrogen deposition

HWp from solid waste this will be included in the next inventory

Co2, CH4 and N2o emissions from Combined Heat and power (CHp) combustion systems

CH4, N2o emissions from biological treatment of waste

Co2, CH4 and N2o emissions from incineration and open burning of waste

IE

Co2, CH4 and N2o emissions from water-borne navigation

Fuel consumption for this source category is included elsewhere. Accurate quantification of fuel consumptionattributed to waterborne navigation was to be undertaken in 2015 and a progress report will be included in the next inventory submission

Co2, CH4 and N2o emissions from off-road vehicles and other machinery

ozone Depleting Substance replacements forfireprotectionandaerosols

1. introduction


NE, IE or NO

Activity Comments

NO

other product manufacture and use

rice cultivation

Co2, CH4 and N2o emissions from Soda ash production

Co2 from Carbon Capture and Storage

Co2, CH4 and N2o emissions from adipic acid production

Co2, CH4 and N2o Caprolactam, Glyoxal and Glyoxylic acid production

precursor emissions have only been estimated for biomass burning, and only for Co and Nox


2.1 Trends for aggregated GHG emissions

this chapter summarizes the trends in GHG emissions during the period 2000 to 2012, by greenhouse gas and by sector. Detailed explanations of these trends are found in Chapters 3 to 6, and a summary table of all emissions for 2012 is provided in Appendix A.

in 2012 the total GHG emissions (excl. FolU) in South africa were estimated at 539 112 Gg Co2eq, a slow increase of 21.7% since 2000 (443 133 Gg Co2eq). the 2000 emissions were also 55.2% higher than the 1990 estimate of 347 346 Gg Co2eq,althoughitisdifficulttodirectly compare the 2000 to the 1990 estimates as the methodologyhaschangedsignificantlyoverthelast20years. the recalculated estimate for 2000 (excl. FolU) is 1.4% lower than the estimate from the previous 2010 inventory (Department of environmental affairs, 2015), which was 449 498 Gg Co2eq (excl. FolU).

Figure 2.1 shows the trends and relative contributions of the different gases to the aggregated national GHG emissions (excl. FolU). there was a 23.9% and 15.1% increase in Co2 and CH4 (in Co2eq), respectively, between 2000 and 2012, and a decline of 5.6% in N2o emissions over that period. Fluorinated gases (F-gases) increased from 982 Gg Co2eq in 2000 to 3 375 Gg Co2eq in 2012, and their contribution to the total GHG emissions increased from 0.2% to 0.6% over the 12 years.

the land and HWp sub-sectors show annual variation, but are estimated to be a net sink of Co2. including the emissions and removals from these two sub-sectors in the overall inventory produces a total GHG emission estimate of 434 304 Gg Co2eq for 2000, which increased by 19.3% to 518 297 Gg Co2eq in 2012 (Figure 2.2). the recalculated total GHG emission estimates for 2000 including the FolU sector are 2.1% higher than the value reported in the National GHG inventory for

2010 (Department of environmental affairs, 2014). this is due to the incorporation of more detailed land maps, improved soil carbon analysis, improved HWp data and the inclusion of other lands. including FolU in the total emissions leads to a slight increase in the Co2 contribution from an average of 84.4% to 83.7% over the 12 years. the contributions from CH4 and N2o increased by 0.5% and 0.2%, respectively; while the average contribution from F-gases remained the same (Figure 2.2).

2. TRENDS IN GHG EMISSIONS

2. trends in GHG emissions


Figure 2.1: Greenhouse gases: trend and emission levels (excl. FolU), 2000 – 2012.

Figure 2.2: Greenhouse gases: trend and emission levels (including FolU sub-sectors), 2000 – 2012.


2.2 Emission trends by gas

2.2.1 Carbon dioxide

Figure 2.3 presents the contribution of the main sectors to the trend in national Co2 emissions (excl. FolU). the energy sector is by far the largest contributor to Co2 emissions in South africa, contributing an average of 91.9% between 2000 and 2012, and 92.5% in 2012. the categories 1A1 energy industries (61.1 %), 1A3 transport

(10.3 %) and 1A4 other sectors (8.8 %) were the major contributors to Co2 emissions in 2012. the ippU sector contribution declined from 8.4% in 2000 to 7.2% in 2012. the aFolU sector (excl. FolU) contributed an average of 0.2% towards the Co2 emission budget over the 12 years.

Figure 2.3: Co2: trend and emission levels of sectors (excl. FolU), 2000 – 2012.



2.2.2 Methane

the sector contributions to the total CH4 emissions (excl. FolU) in South africa are shown in Figure 2.4. National CH4 emissions increased from 47 937 Gg Co2eq (2 084 Gg CH4) in 2000 to 55 168 Gg Co2eq (2 399 Gg CH4) in 2012 (a 15.1% increase). the aFolU livestock category and waste sectors were the major contributors,

providing 50.5% and 38.3%, respectively, to the total CH4 emissions in 2012. the contribution from livestock declined by 11.7%, while the contribution from the waste sector increased by a similar amount (14.1 %) over the period 2000 to 2012.

Figure 2.4: CH4: trend and emission levels of sectors (excl. FolU), 2000 – 2012


2.2.3 Nitrous oxide

Figure 2.5 shows the contribution from the major sectors to the national N2o emissions in South africa. the emissions declined by 5.5% over the 2000 to 2012 period from 27 162 Gg Co2eq (88.4 Gg N2o) to 25 645 Gg Co2eq (86.6 Gg N2o). the category 3C aggregated and non-CO2 sources on land (which includes emissions

from managed soils and biomass burning) contributed an average of 80.5% to the total N2o emissions over the period 2000 to 2012, while the energy sector and livestock subsector (which includes manure management) contributed an average of 8.4% and 4.4%, respectively.

Figure 2.5: N2o: trend and emission levels of sectors, 2000 – 2012.



2.2.4 Fluorinated gases

est imates of Hydrof luorocarbons (HFC) and Perfluorocarbons(PFC)emissionswereonlygivenforthe ippU sector in South africa. emission estimates vary annually between 888 Gg Co2eq and 4 160 Gg Co2eq (Figure 2.6). HFC emissions in 2005 were estimated at 842 Gg Co2eq and these increased to 2 096 Gg Co2eq in 2010 and to 1 396 Gg Co2eq in 2012. there is no data prior to 2005. pFC emissions were estimated at 982 Gg Co2eq in 2000. this increased to 970 Gg Co2eq

in 2007, then declined to 108 Gg Co2eq in 2009 and increased again to 1 979 Gg Co2eq in 2012. there is a sharp decline in emissions from the metal industry between 2006 and 2009 and this is attributed to reduced production caused by electricity supply challenges and decreased demand following the economic crisis that occurred during 2008/2009. increases in 2011 and 2012 were due to increased emissions from aluminium plants duetoinefficientoperations.Theindustrywasusedtocontrol the electricity grid and had to switch on and off at short notice leading to large emissions of C2F4 and CF4.

Figure 2.6: F-Gases: trend and emission levels of sectors, 2000 – 2012.


2.3 Emission trends specified by source category

Figure 2.7 provides an overview of emission trends per ipCC sector in Gg Co2eq (excl. FolU). the energy sector is the largest contributor to the total GHG emissions, providing 77.3% in 2000 and 79.4% in 2012. the aFolU sector (excl. FolU) and the ippU sector contributed 9.6% and 6.9%, respectively, to the total GHG emissions in 2012. their contributions decreased

Figure 2.7: total GHG: trend and levels from sectors (excl. FolU), 2000 – 2012.

by 2.6% and 0.7%, respectively since 2000. the percentage contribution from the waste sector increased from 2.7% to 4.1% over the 12-year period. including the full aFolU sector reduces the total GHG emissions to 518 297 Gg Co2eq (Table 2.1). the contribution from the aFolU sector (incl. FolU) declined from 10.5% in 2000 to 6.0% in 2012. trends in emissions by sub-categories in each sector are described in more detail in Chapters 3 to 6.



table 2.1: trends and levels in GHG emissions for South africa between 2000 and 2012.

Energy IPPUAFOLU

(excl. FOLU)

AFO-LU (incl. FOLU) Waste

Total (excl. FOLU)

Total (incl. FOLU)

(Gg Co2 eq)

2000 342 592 33 564 54 689 45 860 12 288 443 133 434 304

2001 341 338 33 737 54 686 45 813 13 205 442 966 434 093

2002 352 216 35 271 55 378 43 971 14 088 456 953 445 545

2003 375 718 34 759 53 145 42 342 14 947 478 546 467 765

2004 393 607 35 290 53 019 44 066 15 780 497 630 488 744

2005 386 587 38 262 52 830 44 047 16 591 494 270 485 487

2006 393 076 39 298 53 141 41 952 17 381 502 897 491 708

2007 421 753 37 505 53 141 39 758 18 152 529 335 517 168

2008 413 698 34 935 53 297 47 225 18 905 520 633 514 763

2009 421 834 33 026 52 150 39 011 19 639 526 649 513 509

2010 435 117 35 463 53 823 38 456 20 354 544 758 529 391

2011 415 843 38 888 54 087 38 376 21 151 529 969 514 257

2012 428 112 37 129 51 943 31 128 21 929 539 112 518 297

Change 2000 -

2012 (%)24.9 10.6 -5.0 -32.1 78.5 21.7 19.3

table 2.2: precursor GHG: trend and emission levels in Co and Nox (Gg of gas) from biomass burning, 2000 – 2012.

Gas 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Co 1 256 1 449 1 463 1 082 1 027 1 604 1 435 1 251 1 340 1 246 1 524 1 412 1 402

Nox 62 73 73 53 51 79 69 60 65 60 74 68 55

2.4 Emission trends for indirect GHG

the trend in total emissions of carbon monoxide (Co) and nitrogen oxides (Nox) is shown in Table 2.2. these

emissions are recorded only for biomass burning. an average of 1 458 Gg Co and 70 Gg Nox is estimated to have been produced from biomass burning over the period 2000 to 2012. the highest values were recorded in 2005.


3.1 An overview of the energy sector

South africa’s GDp is the 26th highest in the world, but in primary energy consumption South africa is ranked 16th in the world. South africa’s energy intensity is high mainly because the economy is dominated by large-scale, energy-intensiveprimarymineralsbeneficiationindustriesand mining industries. Furthermore, there is a heavy reliance on fossil fuels for the generation of electricity andsignificantproportionoftheliquidfuelsconsumedin the country. the energy sector is critical to the South african economy because it accounts for a total of 15% in the GDp.

in May 2009, the Department of Minerals and energy was divided into two separate departments, namely, the Department of Mineral resources and the Department of energy (Doe). the Doe is responsible for the management, processing, exploration, utilisation and development of South africa’s energy resources.

the Doe’s energy policy is mainly focused on the following key objectives:

• Diversifying primary energy sources and reducing dependency on coal;

• Good governance, which must also facilitate and encourage private-sector investments in the energy sector;

• environmentally responsible energy provision;

• attaining universal access to energy by 2014;

• Achievingafinalenergydemandreductionof12%by 2015; and

• providing accessible, affordable and reliable energy, to the poorer communities of South africa.

the energy sector in South africa is highly dependent on

coal as the main primary energy resource. the largest source of energy sector emissions in South africa is the combustion of fossil fuels. emission products of the combustion process include Co2, N2o, CH4 and H2o. a large quantity of liquid fuels is imported in the form of crude oil. renewable energy comprises biomass and natural processes that can be used as energy sources. Biomass is used commercially in industry to produce process heat and in households for cooking and heating.

the 2004 White paper on renewable energy indicated that the target for renewable energy should be 10 000 GWh by 2013. the Doe recently developed a biofuel strategy to contribute towards the production of renewable energy and to minimize South africa’s reliance on imported crude oil.

3.1.1 Energy demand

in terms of energy demand, South africa is divided into six sectors: industry, agriculture, commerce, residential, transport and other. the industrial sector (which includes mining, iron and steel, chemicals, non-ferrous metals, non-metallic minerals, pulp and paper, food and tobacco, and other) is the largest user of energy in South africa.

3.1.2 Energy reserves and production

the primary energy supply in South africa is dominated by coal (65.7%), followed by crude oil (21.6%), renewable and waste (7.6%) and natural gas (2.8%) (Department of energy, 2010) (Figure 3.1).

3.1.2.1 Coal

Coal provides for about 65% of South africa’s primary energy needs (Department of energy, 2016). South african coal is nearly all bituminous, with very little anthracite. it is generally of low quality with a high ash

3. ENERGY SECTOR

3. energy Sector


content. the contribution of coal to the total primary energy supply decreased by 8% between 2000 and 2006, however this increased by 5% between 2006 and 2009. South africa produces an average of 224 Mt of marketable coalannually,makingitthefifthlargestcoal-producingcountry in the world (GCSi, 2009). South africa has coal reserves of 30 156 Mt, representing 3.5% of total global reserves (SaMi, 2013). in 2012 the coal sector remained the largest revenue earner in the country. local coal sales volume increased by 4.4 percent from 177.9 Mt in 2011 to 185.7 Mt in 2012 (Table 34). electricity consumed 120.4 Mt (64.8 percent) of local sales, followed by the synthetic fuels sector 40.9 Mt (22 percent), industries 8.7 Mt (4.7 percent) and merchants and domestic 8 Mt or 4.3 percent) (SaMi, 2013).

3.1.2.2 Nuclear

Nuclear power contributes a small amount (1.9%) to South africa’s energy supply (Department of energy, 2010). the Koeberg Nuclear power Station, which is the only nuclear power station in africa, has two 900 MW units supplying 1 800 MW to the national grid. it therefore providesasignificantamountofSouthAfrica’selectricity(GCSi, 2009). the National Nuclear regulator, as the main safety regulator responsible for protecting persons, property and the environment against nuclear damage, provides safety standards and regulations. low- and intermediate-level waste from Koeberg is transported by road in steel and concrete containers to a remote disposal site at vaalputs, 600 km away in the Kalahari

Figure 3.1: Sector 1 energy: trend in primary energy consumption in South africa, 2000 – 2012.

7,000,000.00

6,000,000.00

5,000,000.00

4,000,000.00

3,000,000.00

2,000,000.00

1,000,000.00

-

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Coal Gas Hydro Nuclear Crude oil renewable energy

Primary Energy Supply (TJ)


Desert. High-level waste, the spent fuel, is stored on-site in special pools equipped with high-density racking (eskom, 2012). the total consumption of nuclear energy decreased by 1.4% for the period 2000 to 2009.

3.1.2.3 Renewable Energy

renewable energy sources, other than biomass (the energy from plants and plant-derived materials), have not yet been exploited optimally in South africa. renewable energy and waste contribute a total of 7.6% to the energy supply (Department of energy, 2010). Wind as an energy source contributes more than 4 GWh annually. Hydro and geothermal solar contribute a total of 0.2% and 0.1%, respectively, to the primary energy supply. their contribution increased by 0.2% and 2.0%, respectively, between 2000 and 2009.

3.1.2.4 Liquid Fuels

in the third quarter of 2008 the demand for petrol decreased by 10% compared with the same period in 2007, while the demand for diesel increased by 3.0% as industries scaled down operations because of the global economic recession (GCSi, 2009). the petrol price in South africa is linked to international petrol markets in which use the United States dollar. this means the supply and demand of petroleum products on the international markets, combined with the rand-dollar exchange rate influencethedomesticprice.

3.1.2.5 Oil and Gas

South africa has limited crude oil reserves and imports from the Middle east and other african countries to meet 95% of its crude oil requirements. Crude oil consumption increased from 17% in 2000 to 29% in 2002, but then declined to 18% by 2009. limited natural gas reserves exist around the South african coast. the consumption of gas varied between 1% and 3% between 2000 and 2010. refined petroleum products such as petrol, diesel,residualfueloil,paraffin,jetfuel,aviationgasoline,

liquefiedpetroleumgasandrefinerygasareproducedbyrefiningcrudeoil(inoilrefineries),convertingcoalto liquid fuels and gas to liquid fuels, and turning natural gas into liquid fuels. industry is the largest customer of refinedpetroleumproducts.

3.1.2.6 Electricity

South africa’s largest power producer, eskom, generates 95% of the electricity in South africa and about 45% of africa’s electricity (GCSi, 2009). eskom generates, transmits and distributes electricity to industrial, mining, commercial, agricultural and residential customers and redistributors. approximately 88% of South africa’s electricityisgeneratedfromcoal-firedpowerstations,6.5% from Koeberg Nuclear power Station, and 2.3% from hydroelectric and other renewables (GCSi, 2009).

3.1.3 Transport

South africa has roads, rail and air facilities (both domestic and international). in 2010, the South african transport sector employed 392 381 people, representing a total of 0.8% of the population (Statistics Sa, 2011). South africa investedR170billioninthetransportsysteminthefive-year period from 2005/06 to 2009/10, with r13.6 billion of the total allocated to improve public transport systems for the 2010 FiFa World Cup.

3.1.3.1 Rail

State-owned transnet is a freight transport and logistics company. transnet owns and operates the largest rail network on the continent. of the 55 000km of track in sub-Saharan africa, 40% of the operating network and 70%ofthetrafficareoperatedbyTransnet.(Transnetlong-term planning Framework, 2016). the Gautrain rapid rail link commenced operation in 2010. it has 80 km of routes, linking Sandton to or tambo international airport, Johannesburg to pretoria, and rosebank to Johannesburg’s park Station. the aim of the Gautrain was

3. energy Sector


to reduce the congestion along the Johannesburg-pretoria trafficroute,whichaccommodatesapproximately300000 vehicles per day.

3.1.3.2 Road Transport

South africa has the longest road network of any country in africa. the Bus rapid transit (Brt) system implements high-quality public transport networks that operate on exclusive right of way and which incorporates existing bus and minibus operators.

3.1.3.3 Civil Aviation

South africa is home to more than 70% of the aviation activities in the SaDC region. South africa’s aviation industry has experienced a significant growth in the past 10 years. the airports Company of South africa (aCSa) owns and operates the 10 principal airports, including the three major international airports located in Johannesburg, Durban and Cape town.

3.1.3.4 Ports

transnet National ports authority (tNpa) is the largest port authority on the african continent. it owns and manages South africa’s ports. Commercial ports play a crucial role in South africa’s transport, logistics and socio-economic development. approximately 98% of South africa’s exports are conveyed by sea.

3.2 GHG Emissions from the energy sector

the energy sector in South africa is highly dependent on coal as the main primary energy provider. the largest source of energy sector emissions in South africa is the combustion of fossil fuels. emission products of the combustion process include Co2, N2o, CH4 and H2o.

the energy sector includes:

• exploration and exploitation of primary energy sources;

• Conversion of primary energy sources into more useableenergyformsinrefineriesandpowerplants;

• transmission and distribution of fuels; and

• Final use of fuels in stationary and mobile applications.

3.2.1 Overview of shares and trends in emissions

total GHG emissions for the energy sector increased by 25.0% between 2000 and 2012, and produced a total accumulated GHG estimate of 5 121 939 Gg Co2eq over the 12-year period. an analysis of key categories was completedtodeterminethemostsignificantemissionsources in the energy sector. the majority of emissions are from energy industries (63.2 %) (Figure 3.2), followed by 10.7% from transport and 9.1% from manufacturing industries and construction.

the main source of emissions in this sector is Co2 from fossil fuel combustion. the largest source of emissions for the period 2000 to 2012 is the main activity electricity and heat production, which accounted for 54.6% (2 797 150 Gg Co2eq) of the total accumulated emissions from the energy sector. the second largest emitting subcategory was manufacturing industries and construction which accounted for 467 830 Gg Co2eq over the period 2000 to 2012. the transport sector and fugitive emissions from energy production accounted for an accumulated amount of 549 896 Gg Co2eq and 374 087 Gg Co2eq (10.7% and 7.3% of the total emissions) respectively, between 2000 and 2012. the manufacture of solid fuels and other energy industries accounts for an accumulated 391 561 Gg Co2eq of the total GHG emissions in the energy sector for the period 2000 to 2012. the residential and commercial sectors are both heavily reliant on electricity for meeting energy needs, contributing a total of 211 817 Gg Co2eq and 189 211 Gg Co2eq of emissions, respectively.


the total GHG emissions in the energy sector increased from 342 592 GgCo2eq in 2000 to 428 112 Gg Co2eq in 2012 (Figure 3.3). the majority of emissions were from main activity electricity and heat production which accounted for a total of 54.6% of the total GHG emissions from the energy sector in 2012. total GHG emissions increased between 2001 and 2004, mainly because of economic growth and development, which led to increased demand for electricity and fossil fuels. an expansion of industrial production during that period increased the demand for electricity and fossil fuels. economic growth also increased the amount people travelled, leading to higher rates of consumption of petroleum fuels. GHG emissions decreased by 1.78% in 2005, which was followed by an increase in emissions until 2007. the largest annual increase (7.3%) was evident in 2007 (Figure 3.4). the largest decrease was seen in 2011 (4.4%).

Table 3. 1 shows the contribution of the source categories in the energy sector to the total national GHG inventory.

in 1990 and 1994 the energy sector produced an estimated 260 886 Gg Co2eq and 297 564 Gg Co2eq, respectively. Between 1990 and 2000 there was an increase of 31.3% in total GHG emissions from the energy sector, and between 2000 and 2010 there was a further 27.0% increase (Figure 3.5). a small decline (1.6%) is evident between 2010 and 2012. it should be noted, however, that improvements in activity data, emission factors and emission calculations were made between 1990 and this 2000 inventory, therefore some of the increase experienced over this period may be attributed to methodological changes.

Figure 3.2: Sector 1 energy: average contribution of source categories to total energy sector GHG emissions, 2000 and 2012.

3. energy Sector


Figure 3.4: Sector 1 energy: annual change in total GHG emissions between 2000 and 2012.

Figure 3.3: Sector 1 energy: trend and emission levels of source categories, 2000 – 2012.


tabl

e 3.

1:

Sect

or 1

ene

rgy:

Con

trib

utio

n of

the

var

ious

sou

rces

to

tota

l ene

rgy

GH

G e

mis

sion

s.

Tota

l CO

2 (G

g C

o2e

q)

Peri

od20

0020

0120

0220

0320

0420

0520

0620

0720

0820

0920

1020

1120

12

elec

tric

ity g

ener

atio

n18

5 92

518

0 91

518

7 14

920

3 86

621

1 90

520

9 92

021

2 30

423

3 88

322

3 47

722

9 93

723

6 79

823

5 20

524

5 86

5

Petroleumrefining

4 04

93

905

3 39

03

885

3 56

93

418

3 67

53

767

3 87

43

803

3 55

13

341

3 38

4

Man

ufac

ture

of s

olid

fu

els

and

othe

r en

ergy

indu

stri

es30

572

31 0

3630

597

31 0

9531

161

29 4

0328

810

30 3

1229

814

29 8

8329

531

29 2

9430

053

Man

ufac

turi

ng

indu

stri

es a

nd

cons

truc

tion

32 6

5332

182

33 3

8935

899

37 8

7837

148

38 0

7239

463

42 2

7740

129

41 1

1728

413

29 2

12

Dom

estic

avi

atio

n2

047

2 07

92

204

2 62

62

837

3 14

63

118

3 37

43

413

3 46

83

670

3 55

23

478

roa

d tr

ansp

orta

tion

33 3

5433

595

34 0

6835

479

36 8

3437

902

39 0

4641

255

40 1

3040

695

43 4

4044

377

43 8

57

rai

lway

615

604

589

558

582

579

535

552

522

538

496

354

355

Com

mer

cial

/ in

stitu

tiona

l9

549

11 0

3912

201

13 1

7716

155

15 1

9916

486

15 1

5515

388

15 9

7017

114

15 5

8116

196

res

iden

tial

7 10

19

237

11 0

8512

282

13 9

4714

948

16 1

2718

323

20 2

1122

361

24 7

0320

550

20 9

43

agr

icul

ture

/ for

estr

y/

fishing/fishfarms

2 38

72

256

2 32

72

449

2 58

12

665

2 80

93

072

3 02

13

065

3 30

83

430

3 43

2

Nonspecified

989

984

983

1 01

51

045

1 06

21

073

1 10

01

053

1 07

61

139

1 13

81

114

Fugi

tive

emis

sion

s fr

om c

oal m

inin

g 2

003

2 16

11

961

2 11

82

167

2 18

12

180

2 20

52

245

2 23

02

266

2 98

61

760

Fugi

tive

emis

sion

s fr

om n

atur

al g

as

752

753

955

1 45

81

379

1 16

01

133

1 13

31

138

1 24

396

478

664

2

Fugi

tive

emis

sion

s -

othe

r em

issi

ons

from

en

ergy

pro

duct

ion

30 5

9731

039

31 3

1729

812

31 5

6827

854

27 7

0828

160

27 1

3427

435

27 0

1926

836

27 8

21

Tota

l GH

G

emis

sion

s (G

g C

O2e

q)34

2 59

234

1 78

535

2 21

637

5 71

839

3 60

738

6 58

739

3 07

642

1 75

341

3 69

842

1 83

443

5 11

741

5 84

342

8 11

2

3. energy Sector


3.2.1.1 Energy Emissions and the South African Economy

trends in the GHG emissions from fossil fuel combustion aremainlyinfluencedbylong-andshort-termfactorssuch as population increase, economic fluctuations, energy-pricefluctuationsandenergysupplychallenges.GDp performance is the key driver for trends in energy demand in many sectors. population changes also play an importantroleinthefluctuationofGHGemissionsintheresidential sector. in broad terms, energy consumption patterns respond to the changes that affect the scale of consumption, that is, the number of cars, the size ofhouses,population,efficiencyofenergyusageandbehavioural choices/tendencies.

the South african economy is directly related to the global economy, mainly through exports and imports. the GHG emissions in the energy sector have increased by 16.0% over the 12-year period between 2000 and 2012, mainly due to economic growth and development, and preparations for the 2010 FiFa World Cup.

according to Statistics South africa, the growth rate of the economy slowed from 3.6% in 2002 to 1.9% in 2003. the slowdown was attributed to a contraction in the growth of the agricultural and manufacturing sectors. Despite the slowdown in economic growth, GHG emissions still increased by 3.19% in 2002 and 6.67% in 2003in 2004 GDp growth increased again to 3.7% and to 5% in 2005. GHG emissions increased by only 1.8% between 2005 and 2006. the reason for the slower rate of increase in emissions between 2004 and 2005 is unknown.

the annual rate of increase in GHG emissions was significantly lower between 2007 and 2009, with an increase of only 2.0% in 2009. in May 2009 South africa officiallyenteredaneconomicrecession–thefirstin17years – sparked by the global economic recession that started towards the end of 2008. Until then, the country’s economic growth and development had been stable and consistent. according to Statistics South africa, the GDp increased by 2.7% in 2001, 3.7% in 2002, 3.1% in 2003, 4.9% in 2004, 5.0% in 2005, 5.4% in 2006, 5.1% in 2007 and 3.1% in 2008. as a result of the recession, GHG emissions

Figure 3.5: Sector 1 energy: trend and emission levels of total GHG’s from the energy sector, 1990 – 2012.


decreased enormously almost across all categories in the energy sector.

By November 2009, the economy had recovered, achieving a 0.9% growth, primarily because of growth in the manufacturing sector. Hosting the 2010 FiFa World Cup in June and July 2010 also contributed positively to the economy and, as a result, GHG emissions increased by 3.5% that year.

South africa’s economy is highly dependent on a reliable electricity supply. a decline in emissions between November 2007 and the end of January 2008 can be attributed to substantial disruptions in the electricity supply caused by a lack of adequate generating capacity.

in January 2006 the main power producer, eskom, started experiencing difficulties meeting customer demand (NerSa, 2008). this situation deteriorated in late 2007 and early 2008 and eskom resorted to load shedding. the extent of the load shedding had a disruptive impact on businessoperations,traffic,industry,miningoperations,commerce, hospitals, clinics, schools and other institutions, and households. the situation deteriorated to the extent that major mining operations had to close down on 24 January 2008 for safety considerations (NerSa, 2008).

3.2.2 Key sources

level and trend key category analyses were completed for 2012 using 2000 as the base year. the level assessment shows that energy industries (solid fuels) – main activity electricity and heat production (1.a.1.a), road transportation (1.a.3.b), energy industries (liquid fuels) – manufacture of solid fuels and other energy industries (1.a.1.c), manufacturing industries and construction (solid fuels) (1.a.2), and other sectors (solid fuels) – residential (1.a.4.b) are the top subcategories in the energy sector. in the trend assessment, it is other sectors (solid fuels) – residential (1.a.4.b), energy industries (solid fuels) – main activity electricity and heat production (1.a.1.a), manufacturing industries and construction (solid

fuels) (1.a.2), energy industries (liquid fuels) – manufacture of solid fuels and other energy industries (1.a.1.c) and other sectors (liquid fuels) – commercial/institutional that top the list (Table 1.9, Table 1.11). this differs from the 2000 inventory which showed fugitive emissions from coal mining (1B1) to be the second most important emitter, whereas in this inventory these emissions are not seen as a key category. also, emissions from road transportation have move much higher up the key categories list in this inventory.

3.3 Fuel combustion activities [1A]

the combustion of fuels includes both mobile and stationary sources with their respective combustion-related emissions. GHG emissions from the combustion of fossil fuels in this inventory will include the following categories and subcategories:

• 1a1 energy industries

o 1a1a Main activity electricity and heat production

o 1a1b petroleum activity

o 1a1c Manufacture of solid fuels and other energy industries

• 1a2 Manufacturing industries and construction

o 1a2c Chemicals

o 1A2mNon-specifiedsectors

• 1a3transport sector

o 1a3a Civil aviation

o 1a3b road transportation

o 1a3c railways

o 1a3d Waterborne navigation

o 1a3e other transportation

3. energy Sector


• 1a4 other sectors

o 1a4a Commercial/ institutional

o 1a4b residential

o 1A4cAgriculture/forestry/fishing/fishfarms

3.3.1 Comparison of the sectoral approach with the reference approach

the reference approach is a quick estimate of the total Co2emittedinacountryusingafirst-orderestimateofnational GHG emissions based on the energy supplied to a country. the reference approach can be used to estimate a country’s Co2 emissions from fuel combustion and can be compared with the results of the sectoral emission estimates. that was done for this inventory. over the period 2000 to 2012 the Co2 emissions were higher using the reference approach (Appendix D). reporting has improved over the 12-year period and, as a result, the difference between the two approaches has declined from 56% in 2000 to 20% in 2012. there are a number of possible reasons for the discrepancy:

• Netcalorificvalues(NCVs)used inthesectoralapproach differ from those used in the reference approach. in power generation, NCvs in the sectoral approach vary over the 2000–2012 time series based on the information provided by industry;

• activity data on liquid fuels in the sectoral approach, particularly for energy industries, is sourced directly from the companies involved and has been reconciled with other publicly available datasets;

• allocation of solid fuels between energy use, non-energy use as well as use for synfuels production remains one of the key drivers of the differences observed between the two datasets.

a more detailed analysis of the discrepancies between the reference and sectoral approach will be included in the next submission.

3.3.2 Feed stocks and non-energy use of fuels

there are cases where fuels are used as raw materials in production processes. For example, in iron and steel production, coal is used as a feedstock in the manufacture of steel. the 2006 ipCC Guidelines emphasize the significanceofseparatingenergyandprocessemissionsto prevent double counting the industrial and energy sectors. therefore, to avoid double counting, coal used for metallurgical purposes has been accounted for under the ippU sector. information on feed stocks and non-energy use of fuels has been sourced from the national energy balance tables. the sources considered include coal used in iron and steel production, the use of fuels as solvents, lubricants and waxes, and the use of bitumen in road construction.

3.3.3 Energy industries [1A1]

the combustion of fuels by large fuel-extraction and energy-producing industries, electricity producers and petroleumrefineriesarethemainsourcesofemissionsfrom fossil fuels in South africa. the GHG emissions from the manufacture of solid and/or liquid fuels are reported underrefineryemissions.

TheSouthAfricanenergydemandprofilerevealsthattheindustry/manufacturing sector utilizes the largest amount of electricity, followed by the mining, commercial and residential sectors (Department of energy, 2009a). in the event of power disruptions, these sectors are more likely to be impacted. in the case of the manufacturing/industry, mining and commercial sectors, disruptions can result in reduced productivity. Table 3.2 gives a summary of the main electricity users in South africa.


3.3.3.1 Source Category Description

3.3.3.1.1 Main activity electricity and heat Production [1A1a]

Main activity electricity refers to public electricity plants that feed into the national grid and auto electricity producers, which are industrial companies that operate and produce their own electricity. the main energy industries include electricity and heat production, petroleumrefiningandthemanufactureofsolidorliquidfuels. this category includes electricity produced both by public and auto electricity producers.

the energy balances published by the Doe indicate the type of fuel and the quantity consumed. electricity generation is the largest key GHG emission source in South africa, mainly because it mainly uses bituminous coal which is abundantly available in the country. eskom supplies more than 95% of South africa’s electricity needs (Department of energy, 2009). it generates, transmits and distributes electricity to various sectors, such as the industrial, commercial, agricultural and residential sectors. the net maximum electricity generation capacity and electricity consumption for 2000 to 2012 is illustrated in Table 3.3.

additional power stations are being built to meet the

increasing demand for electricity in South africa (eskom, 2011). eskom had planned to invest more than r300 billion in new generation, transmission and distribution capacity up to 2013. in 2008 eskom’s total sales of electricity were estimated at 239 109 GWh. eskom introduced demand side management (DSM) in an effort to reduce electricity consumption by 3 000 MW by March 2011. the utility aims to increase this to 5 000 MW by March 2026. the process involvestheinstallationofenergy-efficienttechnologiesto alter eskom’s load and demand profile. the DSM programme within the residential, commercial and industrial sectors has exponentially grown and exceeded its annual targets. the 2009 saving was 916 MW, against the target of 645 MW. that increased the cumulative saving to 1 999 MW since the inception of DSM in 2008.

3.3.3.1.2Petroleumrefining[1A1b]

this source category includes combustion emissions fromcrudeoilrefiningandexcludesemissionsfromthemanufacture of synthetic fuels from coal and natural gas. Combustion-related emissions from the manufacture of synthetic fuels from coal and natural gas are accounted for under 1a1c. South africa has limited oil reserves and approximately 95% of its crude oil requirements are met byimports.Refinedpetroleumproductssuchaspetrol,diesel,fueloil,paraffin,jetfuelandLPGareproducedby

table 3.2: Sector 1 energy: Summary of electricity users in South africa (Source: Department of energy, 2009a).

Consumer group Electricity consumption Number of consumers

residential 17% 7.5 million

agriculture 3% 103 000

Commercial 13% 255 000

Mining 15% 1100

industry/Manufacturing 38% 33000

transport (mainly railway) 3% 1800

exports 6% 7

own use of distributors 5% not available

Total 100% 7.9 million

3. energy Sector


crudeoilrefining,andtheproductionofcoal-to-liquidfuels and gas-to-liquid fuels.

in 2000 and 2010 the total crude oil distillation capacity ofSouthAfrica’spetroleumrefinerieswas700000bbl/d-

and 703 000 bbl/d, respectively (Sapia, 2006 & 2011). the production of oil was 689 000 tonnes in 2000 and 684 000 tonnes in 2006 (Sapia, 2011). activity data on thefuelconsumedbyrefineriesissourceddirectlyfromrefineries.NationalenergybalancedatafromtheDoEisused to verify data reported by the petroleum industry.

3.3.3.1.3 Manufacture of solid fuels and other energy industries [1A1c]

this category refers to combustion emissions from solid fuels used during the manufacture of secondary and tertiary products, including the production of charcoal. the GHG emissions from the various industrial plants’ own on-site fuel use, and emissions from the combustion of fuels for the generation of electricity and heat for their

table 3.3: Sector 1 energy: Net electricity generation capacity and associated consumption (Source: eskom, 2005, 2007, 2011).

PeriodNet maximum electricity generation

capacity (MW)Total net electricity sold (GWh)

2000 39 810 198 206

2001 39 810 181 511

2002 39 810 187 957

2003 39 810 196 980

2004 39 810 220 152

2005 39 810 256 959

2006 39 810 207 921

2007 37 764 218 120

2008 38 747 224 366

2009 40 506 214 850

2010 40 870 218 591

2011 40 900 224 446

2012 41 500 224 785

own use is also included in this category.

3.3.3.2 Overview of Shares and Trends in Emissions

3.3.3.2.1 Main activity electricity and heat Production [1A1a]

Public electricity producer

the total estimated cumulative coal consumption of the public electricity producer, eskom, for the period 2000 to 2012 was 28 229 590 tJ. Consumption increased by 38.6% over this period (Table 3.4). the total estimated of cumulative GHG emissions from the public electricity producer was 2 736 516 Gg Co2eq between 2000 and 2012.

in 2000, GHG emissions from the public electricity producer totalled 174 702 Gg Co2eq. the main source of emissions in this category was the combustion of coal for electricity generation. the consumption of


electricity increased marginally between 2000 and 2007 (Department of energy, 2009a). the GHG emissions steadily increased throughout the same period from 174 702 Gg Co2eq to 229 218 Gg Co2eq in 2007. the main reason for the increase in GHG emissions during that period was robust economic growth, which increased the demand for electricity and fossil fuel consumption.

During 2003, eskom’s total electricity sales were 196 980 GWh. the peak demand on the integrated system amounted to 31 928 MW and the total GHG emissions during that period were equivalent to 195 Mt Co2eq.ThesefiguresdemonstratethegrowthoftheSouth african economy and the importance of energy as a key driver of the country’s economic growth and development. Between January 2003 and January 2004, South africa increased its electricity output by 7.1%, with a peak demand of 34 195 MW on 13 July 2004, compared with a peak of 31 928 MW in 2003.

table 3.4: Sector 1 energy: Summary of GHG emissions from the public electricity producer.

Period Consumption CO2 CH4 N2O Total GHG

(tJ) (Gg) (Gg Co2 eq) (Gg Co2 eq) (Gg Co2 eq)

2000 1 806 317 173 858 42 802 174 702

2001 1 823 120 175 475 42 809 176 327

2002 1 883 709 181 307 43 836 182 187

2003 2 025 821 194 985 47 899 195 931

2004 2 126 649 204 690 49 944 205 683

2005 2 142 682 206 209 49 951 207 209

2006 2 155 477 207 465 50 957 208 471

2007 2 369 988 228 111 55 1 052 229 218

2008 2 271 791 218 543 53 1 007 219 602

2009 2 335 102 224 579 54 1 034 225 667

2010 2 406 935 231 405 56 1 065 232 526

2011 2 426 965 233 189 57 1 072 234 318

2012 2 537 364 243 497 60 1 118 244 675

in late 2007 and early 2008 the public electricity producer startedtoexperienceddifficultiessupplyingelectricityand resorted to shedding customer loads. the load shedding had a negative impact on the key drivers of economic growth. GHG emissions from the public electricity producer decreased by 4.2% as a result of the electricity disruptions.

the global economic crisis in late 2008 also affected key drivers of growth such as manufacturing and mining sectors. Table 3.2 above provides a breakdown of electricity use by sector as follows: residential 17%, agriculture 3%, Commercial 13%, Mining 15%, industry/ Manufacturing 38%, transport (mainly railway) 3%, exports 6%, and own use of distributors 5% (Department of energy, 2009a).

3. energy Sector


Auto electricity producer

the total estimated accumulated GHG emissions for the period 2000 to 2012 in the auto electricity production category was 60 624 Gg Co2eq. overall, from 2000 to 2012, GHG emissions decreased by 89.4% (Table 3.5). total GHG emissions from auto electricity producers in SouthAfricafluctuatedsignificantlyfromyeartoyear,showing decreases in 2001, 2004, 2005, 2008 and 2011, and increases in the other years. in 2003 the emissions increased by 59.9%. this may be attributed to the economic growth during that period which increased the demand for electricity. the global economic crisis could explain the 16.9% decline in GHG emissions during 2008.


the total cumulative consumption of fuels for petroleum refiningfortheperiod2000to2012wasestimatedat

709696TJ,withrefinerygascontributing76.6%in2010.ThetotalGHGemissionsfrompetroleumrefiningwasestimated at 4 049 Gg Co2eq in 2000, decreasing to 3 384 Gg Co2eq in 2012 (Table 3.6).

In2000refinerygascontributed57.0%tothetotalGHGemissions in this category and this increased to 64.8% in 2012. emissions from residual fuel oil decreased from contributing 16.6% in 2000 to only 9.1% in 2012. a shift fromresidualfueloiltorefinerygasinmostrefineriesis the main driver of emissions reduction in this source category.


total GHG emissions from the manufacture of solid fuels and other energy industries was 30 572 Gg Co2eq in 2000, with a slight decline of 1.7% over the 12-year period

table 3.5: Sector 1 energy: Summary of GHG emissions from auto electricity producers, 2000 – 2012.

Period Consumption CO2 CH4 N2O Total GHG


2000 116 046 11 169 2.67 51.53 11 224

2001 47 346 4 557 1.09 21.02 4 579

2002 51 311 4 939 1.18 22.78 4 963

2003 82 036 7 896 1.89 36.42 7 934

2004 64 333 6 192 1.48 28.56 6 222

2005 28 029 2 698 0.64 12.44 2 711

2006 39 627 3 814 0.91 17.59 3 833

2007 48 233 4 642 1.11 21.46 4 665

2008 40 066 3 856 0.92 17.79 3 875

2009 44 149 4 294 1.02 19.60 4 270

2010 44 171 4 251 1.02 19.61 4 272

2011 9 164 882 0.21 4.07 886

2012 12 305 1 184 0.28 5.46 1 190


Table3.6: Sector1Energy:SummaryofconsumptionoffuelsandGHGemissionsinthepetroleumrefiningcategory(1A1b),2000–2012.

PeriodConsumption CO2 CH4 N2O Total GHG


2000 59 637 4 043 2.27 4.7 4 049

2001 57 599 3 898 2.43 4.65 3 905

2002 50 679 3 385 1.81 3.6 3 390

2003 57 486 3 879 2.12 4.27 3 885

2004 53 292 3 563 1.88 3.70 3 569

2005 51 609 3 413 1.77 3.40 3 418

2006 55 122 3 669 1.96 3.90 3 675

2007 56 073 3 761 2.03 4.04 3 767

2008 57 870 3 868 2.08 4.11 3 874

2009 56 524 3 797 2.07 4.15 3 803

2010 52 519 3 546 1.91 3.83 3 551

2011 50 234 3 336 1.76 3.43 3 341

2012 51 049 3 379 1.77 3.43 3 384

table 3.7: Sector 1 energy: Contribution of Co2, CH4 and N2o to the total emissions from the manufacture of solid fuels and other energy industries category (1a1c), 2000 – 2012

PeriodCO2 CH4 N2O Total GHG

(Gg) (Gg Co2 eq) (Gg Co2 eq) (Gg Co2 eq)

2000 30 454 11.64 105 30 571

2001 30 915 12.85 108 31 036

2002 30 479 11.66 106 30 597

2003 30 970 12.44 112 31 094

2004 31 041 11.82 107 31 160

2005 29 290 11.09 102 29 403

2006 28 699 10.90 100 28 810

2007 30 194 11.64 106 30 312

2008 29 699 11.32 104 29 814

2009 29 767 11.36 105 29 883

2010 29 415 10.95 105 29 531

2011 29 179 10.67 105 29 294

2012 29 903 8.42 142 30 053

3. energy Sector


to 30 053 Gg Co2eq in 2012 (Table 3.7). emissions remained fairly constant over the period 2000 to 2012, with annual changes ranging between -5.6% and 5.2%. Co2 emissions contributed 99.6% to the total GHG emissions from this category.

3.3.3.3 Methodological Issues

3.3.3.3.1 Main activity electricity and heat production [1A1a]

electricity production is the largest source of emissions and, according to the 2006 ipCC Guidelines, it is good practice to use higher-tier methods and emission factors for key categories. as such, Co2 emissions from electricity productionwereestimatedbasedoncountry-specificemissionfactorsandplant-specificactivitydata.Hence,net caloric values (NCvs) reported on an annual basis by the power utility and activity data for each power plant were used in these estimations.


GHGemissionsfrompetroleumrefiningwereestimatedbased on a tier 1 approach and ipCC default emission factors.


a country-specific methodology was applied for the calculation of GHG emissions from the manufacture of solid fuels and energy industries. the GHG emissions from this category were calculated based on actual process material balance analysis.



Data on fuel consumption for public electricity generation was obtained directly from the national power producer

for the period 2000 to 2012. activity data for auto electricity producers was sourced from the energy balance for the period 2000-2006 and extrapolated for the period 2007-2012.Using the technique recommended by the 2006 ipCC Guidelines. to convert fuel quantities into energy units for public electricity generation, the net calorificvaluesestimatedbythenationalutilityannuallywereapplied.Acountry-specificaverageNCVof0.0192tJ/tonne was used to convert fuel quantities into energy units and this was sourced from eskom’s annual report for 2010 (eskom, 2010) for auto electricity producers.

a tier 2 approach with country-specif ic emission factors was used to estimate Co2 emissions from coal consumption (Table 3.8).

For the calculation of CH4 and N2o emissions, a tier 1 approach was used with default emission factors sourced from the 2006 ipCC Guidelines. Furthermore, default factors from these guidelines were applied for the estimation of GHG emissions from other fuels such as other kerosene and diesel oil.


Activitydataforpetroleumrefiningwassourceddirectlyfrompetroleumrefineries.IPCCmethodologieswereappliedforthefillingofdatagapstoensurecompletenessand consistency in the data time series. ipCC 2006 default eFs (ipCC, 2006) were used (Table 3.9).

table 3.8: Sector 1 energy: emission factors for GHG emissions (Source: 2006 ipCC Guidelines, vol 2 and Zhou et al., 2009).

Type of Fuel Emission factor (kg TJ-1)

Co2 CH4 N20

Sub-bituminous coal 96 250 1 1.5

other kerosene 71 500 3 0.6

Gas/ diesel oil 74 100 3 0.6



GHG emissions results for this category were sourced from manufacturing plants (petroSa and Sasol), and thus were calculated based on actual process balance analysis. GHG emissions from charcoal production were not estimated in this category due to a lack of data on fuel use in charcoal production plants, therefore it was assumed that consumption was included under manufacturing and industries (1a2).

3.3.3.5 Uncertainties and Time-series Consistency

according to the ipCC Guidelines, the uncertainties in Co2 emission factors for the combustion of fossil fuels are negligible. the emission factors were determined from the carbon content of the fuel. Uncertainties in CH4 and N2Oemissionfactorswerequitesignificant.the CH4 emission factor has an uncertainty of between 50 and 150%, while the uncertainty on the N2o emission factor can range from one-tenth of the mean value to ten times the mean value. With regards to activity data, statistics of fuel combusted at large sources obtained from direct measurement or obligatory reporting are likely to be within 3% of the central estimate (ipCC, 2006). those default ipCC uncertainty values have been used to report uncertainty for energy industries.

activity data time series for the period 2000 to 2012 were

sourced directly from energy industries. in cases where data gaps were observed, the ipCC methodologies for fillingdatagapswereused.Thatwasmostlythecaseinpetroleumrefining(1A1b)assomerefineriesdidnotrecordfuelconsumptioninthefirstfouryearsofthetimeseries.to ensure consistency in time-series emission estimates, ipCC default emission factors were used for the entire time series for all energy industries. in some cases (e.g. 1a1c), mass balances methods were applied consistently across the time series. the national power utility changed its annual reporting planning cycle from a calendar year toanApril-Marchfinancialyearfrom2006onwards.that affected the time-series consistency, therefore, the national power utility was asked to prepare calendar-year fuel consumption estimates using its monthly fuel consumption statistics.

3.3.3.6 Source-specificQA/QCandVerification

For the quality control of the activity data used to compile emission estimates in energy industries, various publications were used to verify facility-level activity data. the DMr’s SaMi publication was used to verify fuel used for electricity generation (SaMi, 2010). Similarly, a combination of crude oil input reported in the Sapia reports and in the energy balances applied with the ipCC defaultassumptionsonfuelinputinrefinerieswasusedtoverifyfuelconsumptiondatareportedbyrefineries(SAPIA2010). an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvements.

3.3.3.7 Source-specificRecalculations

in the previous 2000 GHG inventory, the activity data was sourced mainly from publicly available publications, such as the national energy balances and Sapia reports. For this inventory report improvements were made by collecting data directly from the national utility and the petroleumrefineries.TheN2o emission factor for coal was corrected. in the previous inventory a value of 3 kg

table 3.9: Sector 1 energy: emission factor for the calculation of GHG emissionsfrompetroleumrefining(Source:2006IPCC Guidelines).

Type of Fuel Emission factor (kg /TJ)

Co2 CH4 N20

residual fuel oil 77 400 3 0.6

petroleum coke 97 500 3 0.6

Refinerygas 57 600 1 0.1

3. energy Sector


tJ-1 was applied but the source of the data could not be traced so the default ipCC emission factor was applied. as more accurate activity data was collected for all sources and emission factors updated, the emissions within the energy industries sector had to be recalculated back to 2000. the recalculation improved the accuracy of the 2000 GHG emissions result for the energy sector.

3.3.3.8 Source-specificPlannedImprovements and Recommendations


the electricity generation sector is a key category and its estimatehasasignificantinfluenceonthecountry’stotalinventory of GHGs. therefore increasing the accuracy of GHGcalculationsbyapplyingcountry-specificemissionfactors for this sector will improve the national GHG inventory estimate. other improvements for this category would be to:

• formalise the data collection process to ensure continuous collection of data and time-series consistency;

• Collectplantspecificdataforcoalcombusted;

• obtain more detailed information from the national power producer to assist in the explanation of trends throughout the reporting period;

• obtain a list of auto power producers and obtain data directly from the producers. this is important going forward since growth is expected within this sector.

Finally, for purposes of completion, future inventories should consider the collection and reporting of SF6 for thiscategory.Inadditiontotheimprovementsidentifiedabove, the Dea is in the process of promulgating legislation - the National GHG reporting regulations. the purpose of these regulations is to introduce a single national reporting system for the transparent reporting of greenhouse gas emissions, which will be used -

• to update and maintain a National Greenhouse Gas inventory;

• for the republic of South africa to meet its reporting obligations under the United Framework Convention on Climate Change and instrument treaties to which it is bound; and

• to inform the formulation and implementation of legislation and policy.

Data providers operating within the energy, ippU and Waste sectors will be required to regularly report emissions data to Dea on an annual basis, provided that the reporting thresholds are met. it is envisaged that the implementation of these reporting regulations will enhance accuracy, completeness and transparency moving forward.


to improve the reporting of GHG emissions in this categoryitisimportantthatthepetroleumrefineriesprovideplant-specificactivitydata,suchasnetcaloricandcarboncontentvalues,andalsodevelopcountry-specificemission factors that can be used for the calculation of GHG emissions.


to improve the estimation of GHG emissions from the manufacture of solid fuels and energy industries, a more regular collection of activity data would be useful. that would improve the time series and consistency of the data. another improvement would be to monitor the causeoffluctuationsinthemanufacture of solid fuels and other energy industries regularly, to enable the inventory compilerstoelaborateonthefluctuations.


3.3.4 Manufacturing industries and construction [1A2]


according to the 2006 ipCC Guidelines, this category comprise a variety of fuel combustion emission sources, mainly in the industrial sector. in manufacturing industries, raw materials are converted into products using fuels as the main source of energy. the industrial sector consumes 40.8%ofthefinalenergysuppliedinSouthAfrica.Themanufacturing industries and construction subsector can be divided into mining, iron and steel, chemicals, non-ferrous metals, non-metallic minerals, pulp and paper, food and tobacco and other productions (includes manufacturing, construction, textiles, wood products etc.) categories. the largest category is iron and steel which consumes 27.4% of the total energy utilized by the industrial sector (Department of energy, 2009a). emissions from the combustion of fossil fuels in the construction sector are also included in this category. according to the energy

balances compiled by the Doe, fossil fuels used in the construction sector include lpG, gas/diesel oil, residual fuel oil, other kerosene, bitumen, sub-bituminous coal and natural gas.


the estimated cumulative total GHG emissions in the category manufacturing industries and construction for the 12-year period was 467 832 Gg Co2eq. GHG emissions were estimated to be 32 653 Gg Co2eq in 2000 and these decreased by 10.5% to 29 212 Gg Co2eq in 2012 (Figure 3.6).

Solid fuels contributed 86.0% to the total in 2010, while liquid and gaseous fuels contributed 4.6% and 9.3%, respectively. there has been a 2.5% increase in the contribution from gaseous fuels, a 1% increase from liquid fuels and a 3.5% decline in the contribution from solid fuels over the 12-year period.

Figure 3.6: Sector 1 energy: trend in sources of total GHG emissions from fuel used in the manufacturing industries and construction category (1a2), 2000 - 2012.

3. energy Sector


GHG emissions from this category increased by 7.5% in 2003, but decreased by 1.9% in 2005. in 2009 GHG emissions from this category decreased by 5.1%, which might have been a result of the global economic crisis that started in late 2008.

the real value added by the construction sector increased at an annual rate of 10.6% in the second quarter of 2008, lowerthantherateof14.9%recordedinthefirstquarterof2008.Thatreducedgrowthreflecteddeterioratingconditions in the residential and non-residential building sectors, as developers experienced a strain of higher interestratesandescalatinginflationarypressures.


GHG emissions included in this subsector are mainly from fuel combusted for heating purposes. Fuels used as feed stocks and other non-energy uses are accounted for

under the ippU sector. For the manufacturing industries and construction subsector, a tier 1 methodology was applied by multiplying activity data (fuel consumed) with ipCC default emission factors. in future, facility-level data hastobesourcedandcountry-specificemissionfactorsneed to be developed in order to move towards a tier 2 methodology.


For the manufacturing industries and construction sector data for solid fuels for the period 2000 to 2007 were sourced from the Doe’s energy digest, for the period 2007 to 2010 the SaMi report was used to extrapolate the fuel consumption. the activity data on liquid fuels for this category was sourced from Sapia. Data from industries werealsoacquiredandusedtocomparethefiguresintheenergy digest and the SaMi report. Table 3.10 shows the total fuel consumption in this category for the period

table 3.10: Sector 1 energy: Fuel consumption (tJ) in the manufacturing industries and construction category, 2000 – 2012.

PeriodOther

Kerosene (tJ)

Gas/Diesel

Oil (tJ)

Residual Fuel Oil

(tJ)

LPG (tJ)

bitumen(tJ)

Sub-bitu-minous Coal (tJ)

Natural Gas (tJ)

Total(tJ)

2000 698 9 531 194 109 5 053 302 354 39 532 357 471

2001 640 9 888 194 115 5 584 295 804 41 241 353 465

2002 606 10 410 187 113 6 161 306 401 43 048 366 927

2003 626 11 069 185 107 6 276 328 424 48 749 395 436

2004 649 11 702 199 108 6 382 347 344 50 361 416 745

2005 619 12 367 171 106 7 038 337 162 53 166 410 629

2006 601 13 271 166 116 7 245 344 183 56 038 421 621

2007 567 14 870 164 122 7 707 355 304 58 908 437 643

2008 433 14 877 164 118 7 475 383 032 61 778 467 877

2009 444 14 877 207 105 7 602 359 011 64 645 446 892

2010 469 16 129 219 111 8 044 365 687 68 406 459 066

2011 473 17 107 167 138 7 536 243 904 51 382 320 707

2012 383 17 163 198 126 9 807 254 262 44 518 326 458


2000 to 2012. an NCv of 0.0243 tJ/tonne was used to convert fuel quantities into energy units (Department of energy, 2009a). to avoid double counting fuel activity data, the fuel consumption associated with petroleum refining(1A1b)wassubtractedfromthefuelconsumptionactivity data sourced for 1a2.

3.3.4.4.1 Emission factors

Acountry-specificemissionfactors forCO2 for sub-bituminous coal was applied (see section 3.3.3.4). this Country-SpecificEFwasusedforallyearsgoingbackto2000. For all other fuels the ipCC 2006 default emission factors were used to estimate emissions from the manufacturing industries and construction sector (Table 3.11). the default eF’s were applicable to all activities within this sector, since similar fuels were combusted.

3.3.4.5 Uncertainty and Time-series Consistency

according to the 2006 ipCC Guidelines, uncertainty associated with default emission factors for industrial combustion is as high as 7% for Co2, ranges from 50 to 150% for CH4 and is an order of magnitude for N2o. Uncertainty associated with activity data based on less-

developed statistical systems was in the range of 10 to 15%. to ensure time-series consistency in this source category the same emission factors were used for the complete time-series estimates. activity data sourced on fuel consumption was complete and hence there was no needtoapplyIPCCmethodologiesforfillingdatagaps.


the national energy balances and the digest of energy statistics were used to verify fuel consumption data reported in the SaMi report. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of the inventory and identify areas of improvements.


in the previous 2000 GHG inventory, activity data for this category were sourced from the energy balances published by the Doe. For this inventory a combination of the energy digest and the SaMi report was used as a source of activity data for solid fuels, and Sapia was the source of activity data from liquid fuels. the combination of a variety of publicly available information has improved the accuracy of the activity data for this category. Country specific Co2 emissions were also included for coal consumption. the N2o emission factor for coal was corrected. in the previous inventory a value of 3 kg tJ-1 was applied but the source of the data could not be traced so the default ipCC emission factor was applied. recalculations were therefore completed for all years between 2000 and 2012.


in future, facility-level data needs to be sourced and country-specificemissionfactorshavetobedevelopedin order to move towards a tier 2 methodology. the reliance on energy balances and other publications for the

table 3.11: Sector 1 energy: emission factors used in the manufacturing industries and construction category (Source: 2006 ipCC Guidelines).

Type of Fuel Emission factor (kg tJ-1)

Co2 CH4 N20




Liquefiedpetroleumgas (lpG)

63 100 1 0.1

Natural gas (dry) 56 100 1 0.1


Bitumen 80 700 3 0.6

3. energy Sector


compilation of emissions needs to be reduced by sourcing facility-level activity data. this would help to reduce the uncertainty associated with the activity data.

3.3.5 Transport [1A3]

according to the 2006 ipCC Guidelines, estimates of GHG emissions from mobile combustion refers to major transport activities such as road, off-road, air, railways and waterborne navigation. this category only includes direct emissions from transport activities, mainly from liquid fuels (gasoline, diesel, aviation gas and jet fuel). Secondary fuels, such as electricity used by trains, are reported under the main activity electricity and heat production category and not under the transport category. the diversity of sources and combustion takes into consideration the age ofthefleet,maintenance,thesulphurcontentofthefuelused and patterns of use of the various transport modes. the GHG inventory includes emissions from combustion and evaporation of fuels for all transport activity.

the 2006 ipCC Guidelines indicate that, where possible, activitiessuchasagriculturalmachinery,fishingboatsandmilitary transport should be recorded separately under the appropriate sectors and not in the transport sector (ipCC 2006, p.3.8). Furthermore, GHG emissions from fuels sold to any air or marine vessels engaged in international transport are excluded from the national total emissions and are reported separately under the memo items.


3.3.5.1.1 Civil aviation [1A3a]

Civil aviation emissions are produced from the combustion of jet fuel (jet kerosene and jet gasoline) and aviation gasoline. aircraft engine emissions (ground emissions and cruise emissions) are roughly composed of 70% Co2, less than 30% water and 1.0% of other components (Nox, Co, Sox, NMvoCs, particulates, and trace components). Civil aviation data were sourced from both domestic and international aircrafts, including departures and arrivals.

that also included civil commercial use of airplanes, scheduledandchartertrafficforpassengersandfreight,airtaxing, agricultural airplanes, private jets and helicopters. the GHG emissions from military aviation are reported separately under the other category or the memo item multilateral operations.

International aviation (international bunkers) [1A3aI]

GHG emissions from aircraft that returned from an international destination or were going to an international airport were included under this sub-category. that included civil commercial use of airplanes, scheduled andchartertrafficforpassengersandfreight,airtaxiing,agricultural airplanes, private jets and helicopters. the GHG emissions from military aviation were reported separately under the other category or under the memo item multilateral operations.

3.3.5.1.2Roadtransport[1A3b]

according to the 2006 ipCC Guidelines, road transport emissions include fuel consumption by light-duty vehicles (cars and light delivery vehicles), heavy-duty vehicles (trucks, buses and tractors) and motorcycles (including mopeds, scooters and three-wheelers). Fuels used by agricultural vehicles on paved roads are also included in this category. the Doe’s energy balances list the fuels under road transport as diesel, gasoline, other kerosene, residual fuel oil and lpG. road transport was responsible for the largest fuel consumed in the transport sector (79.1% in 2012). Motor gas contributed 64.9% of the road transport fuel consumption in 2010, followed by gas/diesel oil (35%). Between 2000 and 2012 there was an increase in the percentage contribution of gas/diesel oil to road transport consumption (80.1%), and a corresponding decline of 12.7% in the contribution from motor gasoline (Figure 3.7).Thiscanbeattributedtotheefficiencyand affordability of diesel compared with motor gasoline.


3.3.5.1.3 Railways [1A3c]

railway locomotives are mostly one of three types: diesel, electric or steam. Diesel locomotives generally use engines in combination with a generator to produce the energy required to power the locomotive. electric locomotives are powered by electricity generated at power stations and other sources. Steam locomotives are generally used for local operations, primarily as tourist attractions and their GHG emissions are very low (DMe, 2002).Bothfreightandpassengerrailwaytrafficgeneratesemissions. South africa’s railway sector uses electricity as its main source of energy, with diesel being the only other energy source (DMe, 2002).

3.3.5.1.4Waterbornenavigation[1A3d]

according to the 2006 ipCC Guidelines, waterborne navigation sources include emissions from the use of fossil fuels in all waterborne transport, from recreational craft

tolargeoceancargoships,butexcludefishingvessels.Fishing vessels are accounted for under the other sector, inthefishingsubcategory.Thevesselsaredrivenprimarilyby large, medium to slow diesel engines and sometimes by steam or gas turbines.

International waterborne navigation (international bunkers) [1A3dI]

international waterborne navigation GHG emissions includefuelsusedbyvesselsofallflagsthatareengagedin international waterborne navigation. international navigation may take place at sea, on inland lakes and waterways, and in coastal waters. according to the 2006 ipCC Guidelines (p. 3.86), it includes GHG emissions from journeys that depart in one country and arrive in adifferentcountry,andexcludesconsumptionbyfishingvessels. international waterborne navigation was not estimated in this inventory due to a lack of data. as a

Figure 3.7: Sector 1 energy: percentage contribution of the various fuel types to fuel consumption in the road transport category (1a3b), 2000 – 2012.

lpG

residual Fuel oil

Gas / Diesel oil

other Kerosene

Motor Gasoline

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%

3. energy Sector


result, fuel consumption for marine bunkers was included in the national totals. this is not consistent with the 2006 ipCC Guidelines, which require marine bunkers to be reported separately from the national totals. in the future, improved data on marine activities will assist in improving the accuracy of emissions estimates for both waterborne navigation and marine bunkers.

3.3.5.2 Overview of Shares and Trends in Emission

it was estimated that for the period 2000 to 2012, the total cumulative GHG emissions from transport activities was 549 923 Gg Co2eq. GHG emissions from transport activities have increased by 32.4%, from 36 016 Gg Co2eq in 2000 to 47 690 Gg Co2eq in 2012 (Figure 3.8). Co2 emissions from all modes contributed the most to the GHG emissions, while the CH4 and N2o emission

contributions were relatively small (Figure 3.9).

road transport contributed 92.0% towards the total transport GHG emissions in 2012 (43 857 Gg Co2 eq), while 7.3% was from domestic civil aviation and less than 1% from railways. the increase in GHG emissions in the transport sector can be attributed to the increase in motor vehicle from 4.2% in 2000 to 15.7% in 2010 (Stats Sa, 2011). in 2008 vehicles sold amounted to 34 400, which was 16.5% lower than the total units sold in 2007 (Stats Sa, 2007), hence the 2.7% decrease in emissions. this decrease was linked to the global economic crisis. Motor vehicle sales decreased by a further 10.5% in 2009; however in November 2009 the economy started to recover, achieving growth of 0.9%. this growth was accompanied by an increase in GHG emissions by 1.4%.

Figure 3.8: Sector 1 energy: trend in total GHG emissions from the transport sector, 2000 – 2012.


3.3.5.2.1 Transport emissions and the economy

in the transport sector 92% of the GHG emissions were generated from road transport. there is a strong linkage between vehicle population and energy demand. it was estimated that the purchase of motor vehicles increased from 4.81% in 2000 to 10.23% in 2006 (Statistics Sa, 2007).

energy fuels from transport activities consist mainly of liquid fuels. the most dominant fuel is motor gasoline (53.3%), followed by diesel (34%) and then jet fuel (10.9%). the demand for petrol and diesel has remained relatively static over the years. the demand for jet fuel has grown steadily, however, as a result of increased business and tourism activities. in 2001, total liquid-fuel sales grew by 0.3% to 20 934 million litres (Ml). these figures demonstrate the growth of the South african economy and the importance of energy as a key driver of the economy.

the primary driver for the transport sector is GDp growth. For road passenger travel, GDp growth means an increase in the number of commuters and an increase in personal wealth, both in terms of the number of wealthier people and the size of people’s expendable incomes. this results in more money being available to purchase cars and for leisure activities, which, in turn, increases the demand for transport and transport fuel.

in terms of civil aviation, there has been an in increase in the number of passengers who disembarked from internationalscheduledflightsinpastyears.In2008thetotal number of passengers decreased by 7.3% (7.8 million) inthe08/09financialyearcomparedtothe07/08financialyear. However, passenger activity rose by 5% in the 10/11 financialyearamountingto8.2millionpassengers(ACSA,2013).

passengers travelling by commuter rail increased to 646.2 millioninthefinancialyear2008/09,a9%increaseon

Figure 3.9: Sector 1 energy: percentage contribution of Co2, CH4 and N2o from the transport sector, average over 2000 – 2012 period.

3. energy Sector


2008/09 (Stats Sa, 2007). passenger kilometres travelled by trains also increased by 9% to 16.9 billion kilometres in 2008/09.

annual growth in GHG emissions from the transport sector slowed between 2003 and 2004, from 4.9% to 4.1%. this was mainly because the price of oil was at the narrow range of approximately $22/28/bbl, however intheyear2005thepriceofoilescalatedsignificantly.

according to Sapia, the factors that influenced the escalating crude oil price in 2005 were the following:

• Feared shortages due to limited surplus crude oil productionandrefiningcapacityatatimeofstronglyrising world demand for petroleum products, notably in the United States and China;

• a particular shortage of sweet (low-sulphur) crudes duetoalackofrefiningcapacitytoprocesssour(high-sulphur) crudes into the requisite product qualities needed by world markets;

• political tensions in certain crude oil-producing countries; and

• the petrol price in South africa is linked to the price of petrol in US dollars and to certain other international petrol markets.

the local prices of petroleum products are affected by the Zar-USD exchange rate and the dollar price of crude oil. it is important to note that in late 2001 when the exchange rate rose above r12 to the dollar and the oil price was about $20/bbl, crude oil cost some r240/bbl, and in 2004 when the rand strengthened to r6 to the dollar and the crude oil price was $40/bbl, oil still cost someR240/bbl.Thatmeant,despitesomefluctuations,the rand price of crude oil was relatively stable until the dollar price increased above $40/bbl. at a price of $65/bbl and at a rate of r6, 5 to the dollar, the cost became R423/bbl,averysignificantrise.

in 2007, aggregate sales of major petroleum products showedastrongincreaseof7.3%inthefirstquarter

compared with the first quarter of 2006. the most significant increases were in diesel (13.1%), bitumen (36.3%) and lpG (15%). petrol sales grew by 4.4% and jetfuelsalesby4.6%.Paraffinsalesdeclinedby13.4%,indicating that the product was being used less frequently forhouseholdenergy.Inthefirstquarterof2007,thepercentage split of petrol sales between unleaded petrol (Ulp) and lead replacement petrol (lrp) was 64% and 36%, respectively.

in 2008 GHG emissions from the transport sector decreased by 2.5%, which was attributed to the global economic crisis that occurred in 2008 and 2009. total sales of major petroleum products showed an increase of4%inthefirstquarterof2008comparedwiththefirstquarterof2007.Themostsignificantincreaseswereindiesel (9.5%) and industrial heating fuels (35.6%). petrol andparaffinsalesdeclinedby0.9%and3.2%,respectively,affected by price increases, while jet fuel sales grew by 3.3%. lpG volumes were the same as in 2007 and bitumen volumes increased by 7.6%.

in November 2009 South africa’s economy recovered from the recession and in 2010 the country successfully hosted the FiFa World Cup.


it was estimated that for the period 2000 to 2012 the total cumulative GHG emissions from international aviation activities was 33 075 Gg Co2eq. Table 3.12 provides a summary of GHG emissions for the period 2000 to 2012 from international aviation. over the 12-year assessment period, GHG emissions emanating from international aviation have decreased by 18.8%. However in 2006 and 2010 GHG emissions increased by 10.7% and 5.8%, respectively. the increase in 2010 could be attributed to the fact that South africa hosted the FiFa World Cup and international travel increased as a result.



NCvs, which were applied in the transport sector to convert fuel quantities into energy units, were sourced from the Doe (Table 3.13).

3.3.5.3.1 Civil Aviation [1A3a]

the main challenge in this category was splitting the fuel consumptionofinternationalanddomesticflights.The2006 ipCC Guidelines (p.3.78) propose that international/domestic splits should be determined on the basis of departureandlandinglocationsforeachflightstageandnot by the nationality of the airline. the energy balances published by the Doe note that splits for international/national have been made, but the methodology is not described. Furthermore, the energy balances do not give details on whether military aviation activities are included, but,thismaybeduetoconfidentialityissues.TheTier1methodology was used for to calculate aviation emissions as operational data was not available. the tier 1 approach makes use of consumption of fuel and fuel emission factors.


the 2006 ipCC Guidelines suggest that the fuel

table 3.12: Sector 1 energy: Summary of GHG emissions from international aviation (international bunkers), 2000 – 2012.

PeriodCO2 CH4 N2O Total GHG

(Gg) (Gg Co2 eq) (Gg Co2 eq) (Gg Co2 eq)

2000 2 972 3.12 7.43 2 983

2001 2 708 2.84 6.77 2 718

2002 2 687 2.82 6.72 2 697

2003 2 584 2.71 6.46 2 593

2004 2 316 2.43 5.79 2 324

2005 2 267 2.38 5.67 2 275

2006 2 510 2.63 6.28 2 519

2007 2 557 2.68 6.39 2 566

2008 2 478 2.60 6.20 2 487

2009 2 423 2.54 6.06 2 431

2010 2 564 2.69 6.41 2 573

2011 2 482 2.60 6.21 2 491

2012 2 414 2.53 6.04 2 422

Table3.13:NetCalorificValuesforthetransportSector(Source: Department of energy 2009a)

Type of Fuel Net Caloric Value (tJ/l)

Motor gasoline 0.0000342

Gas/ diesel oil 0.0000381

residual fuel oil 0.0000416

aviation gasoline 0.0000343

Bitumen 0.0000402

lpG 0.0000267

Bitumen 80 700

3. energy Sector


consumption approach is appropriate for Co2 emissions as it depends entirely on the carbon content of the fuel combusted, whereas the kilometre approach (distance travelled by vehicle type) is appropriate for CH4 and N2o. Hence, in order to use a higher-tier approach for calculating road transport emissions, a better understanding of the fuel sold and vehicle kilometres travelled is required for theentireSouthAfricanvehiclefleet.Thisdatawasnotavailable for the entire 12-year period, therefore, a tier 1 approach based on fuel consumption and 2006 ipCC emission factors was used to calculate the emissions.


the tier 1 approach was used for the calculation of railway emissions. Default emissions factors from the 2006 ipCC Guidelines were used. the use of a higher-tier approach depends on the availability of fuel consumption data by locomotivetypeand/orcountry-specificemissionfactors.



activity data on fuel consumption was sourced from Sapia’s annual reports (Table 3.14). the 2006 ipCC Guidelines (p. 3.78) require only domestic aviation to be included in the national totals. Hence, in order to separate international from domestic aviation, the Doe energy balances were used to estimate the ratio of domestic to international consumption. Furthermore, according to the 2006 ipCC Guidelines, it is good practice to separate military aviation from domestic aviation. However, based on the Sapia data and national energy balances, it is not possible to estimate the amount of fuel used for military aviation activities. it is not indicated in data sources, therefore, military aviation emissions are thought to be accounted for under domestic aviation. in the D0e’s energy balances civil aviation fuels include gasworks gas, aviation gasoline and jet kerosene.


the energy balances published by the Doe were the main source of data for the amount of fuel consumed (Table 3.14). they did not indicate whether aviation fuel consumptionfiguresincludedmilitaryaviationactivities(these might have been excluded for security reasons).


Sapia annual reports were the main sources of activity data for the transport sector (Table 3.15). the Sapia report on the impact of liquid fuels on air pollution was used to disaggregate fuel consumption into the various users (Sapia, 2008). Where possible, the Doe energy balances where used to verify activity data even though theydonotprovidesufficientinformationforaproperunderstanding of fuel consumption.


the national railway operator, transnet, provided activity data on diesel fuel consumption for the national railway fleet.TheSAPIAreportontheimpactof liquidfuelson air pollution (Sapia, 2008) was used to disaggregate actual diesel consumption for the rail transport sector. an assumption was made that the split of diesel consumption for rail activities was constant for the whole time series (2000 – 2012), which may not necessarily be accurate. to improve accuracy in the future, data should be collected at the subcategory level where annual variations in the activity data can be sourced.

3.3.5.4.4Water-bornenavigation[1A3d]

Lackof source-specificactivitydatamade itdifficultto separately estimate emissions for this sub-category. Heavy fuel oil (HFo) consumption as reported in the Sapia annual reports, which is used for activities in this sub-category, was accounted for under the industrial, commercial and residential sub-category in the other


table 3.14: Sector 1 energy: Fuel consumption (tJ) in the transport sector, 2000 – 2012.

Domestic Aviation Road Transportation Railways

Aviation Gasoline

Jet Kerosene

Motor Gasoline

Other Kerosene

Gas/Diesel

Oil

Residual Fuel Oil LPG

Gas/Diesel

Oil

Period (tJ) (tJ) (tJ) (tJ) (tJ) (tJ) (tJ) (tJ)

2000 835 27 714 337 766 316 123 904 113 54 7 442

2001 880 28 113 335 947 289 128 540 114 0 7 307

2002 843 29 888 335 784 274 135 336 109 0 7 123

2003 764 35 854 346 571 283 143 895 108 0 6 749

2004 760 38 803 356 889 294 152 129 116 0 7 043

2005 802 43 070 362 751 280 160 774 100 54 7 009

2006 745 42 727 366 455 272 172 523 97 0 6 467

2007 758 46 286 375 519 257 193 306 96 0 6 672

2008 752 46 840 359 632 196 193 405 96 0 6 317

2009 746 47 613 367 559 201 193 405 121 0 6 504

2010 790 50 383 388 942 201 209 678 128 0 6 006

2011 745 48 775 388 678 214 222 390 97 0 5 514.30

2012 1070 47 433 380 588 173 223 123 116 0 5638.72

table 3.15: Sector 1 energy: Fuel consumption (tJ) in the international aviation category, 2000 – 2012.

Period Jet Kerosene (tJ)

2000 41 572

2001 37 880

2002 37 580

2003 36 142

2004 32 395

2005 31 704

2006 35 100

2007 35 760

2008 34 657

2009 33 884

2010 35 855

2011 34 711

2012 33 755

3. energy Sector


subsector. as a result, emissions from this sub-category are “implied elsewhere” (ie). Hence, to improve transparency in reporting and the accuracy of emission estimates in the future activity data needs to be further disaggregated to the sub-category level.


ipCC 2006 default emission factors were used to estimate GHG emissions from the transport sector (Table 3.16). the GHG emission factors are applicable to all activities within this sector since similar fuels are combusted.



according to the 2006 ipCC Guidelines, the uncertainty onemission factorsmaybesignificant.Fornon-CO2 emission factors the uncertainty ranges between -57% to +100% and for Co2 emission factors it is approximately 5%, as they are dependent on the carbon content of the fuel and the fraction oxidized (ipCC, 2006, p.3.65).


according to the 2006 ipCC Guidelines, the uncertainties in emission factors for CH4 and N2o are relatively high and are likely to be a factor of 2 to 3%. they also depend onthefollowing:fleetagedistribution;uncertaintiesinthe maintenance pattern of vehicle stock; uncertainties related to combustion conditions and driving patterns; and application rates of post-emission control technologies (e.g. three-way catalytic converters), to mention a few.

activity data were another primary source of uncertainty in the emission estimates. according to the ipCC Guidelines, possible sources of uncertainty, are typically +/-5% due to the following: uncertainties in national energy sources of data; unrecorded cross-border transfers;misclassificationoffuels;misclassificationofvehicle stocks; lack of completeness; and uncertainty in conversion factors from one set of activity data to another.


the GHG emissions from railways or locomotives are typically smaller than those from road transport because less fuel is consumed. also operations often occur on electrifiedlines,inwhichcasetheemissionsassociatedwith railway energy use will be reported under power generation and will depend on the characteristics of that sector. according to the ipCC Guidelines, possible sources of uncertainty are typically +/-5% due to uncertainties in national energy sources of data; unrecorded cross-border transfers;misclassificationoffuels;misclassificationofvehicle stocks; lack of completeness and uncertainty in conversion factors from one set of activity data to another.


according to the ipCC Guidelines, Co2 emission factors for fuel are generally well determined because of their

table 3.16: Sector 1 energy: emission factors used for transport sector emission calculations (Source: 2006 ipCC Guidelines).


Co2 CH4 N20

Motor gasoline 69 300 33 3.2


Gas/ diesel oil 74 100 3.9 3.9

Gas/ diesel oil (railways)

74 100 4.15 28.6


aviation gasoline 70 000 3 0.6

Jet kerosene 71 500 3 0.6


dependence on the carbon content. therefore, the uncertainties around waterborne navigation emission estimates are related to the difficulty distinguishing between domestic and international fuel consumption. With complete survey data, the uncertainty may be as low +/- 5%, while for estimates or incomplete surveys the uncertainties maybe be high as -50%.


Nosource-specificQA/QCandverificationstepsweretaken for this source category.


in the 2000 GHG inventory, the activity data for this category was sourced from the energy balances published by the Doe. For this inventory (2000 – 2012) the energy balances were used as a source of activity data, but the Sapia report on the impact of liquid fuels on air pollution (Sapia, 2008) was used to disaggregate fuel consumption into the various users. this resulted in the recalculation of GHG emissions for these years so as to reduce the uncertainty of the emission estimates.


this category is a key category and it is essential that further work is done to move toward the use of a higher tier emissions estimation methodology.


the road transport sector is a key category and its estimatehasasignificantinfluenceonthecountry’stotalinventory of GHGs. therefore increasing the accuracy of GHG calculations by using bottom-up method and applyingcountry-specificemissionfactorsforthissectorwill improve the national GHG inventory estimate. planned improvements for road transportation include:

Collection of information on vehicle Kilometres travelled (vKt) in the transport sector and to development of a modelling tool for the transport sector. Collecting information on vKt will improve the estimation accuracy of emissions from this category. this project will ensure that emissions estimated from the transport sector are more accurate and will involve (a) bottom-up fuel consumption and vKt for road transportation by vehicle type, (b) bottom-up fuel consumption by vehicle type and (c) a model to ensure sustainability of this work.

Developmentofcountry-specificRoadTransportEmissionFactors for Co2, CH4 and N2o. this study will enable estimation and compilation of the Greenhouse Gas emissions inventory for road transportation using country-specificemissionfactors,thusimprovingtheaccuracyinthe estimation of emissions from this category by moving to a tier 2 methodology. the implementation of this project will involve (a) Detailed measurement of the carbon content of Co2 for petrol and Diesel; and (b) measurement of CH4 and N2o emission factors by vehicle type.


improvement of emission estimation for this category requires an understanding of aviation parameters, including the number of landings/take-offs (ltos), fuel use and the approaches used to distinguish between domestic/internationalflights.Thiswouldensuretheuseof higher-tier approaches for the estimation of emissions. to improve transparency of reporting, military aviation should be removed from domestic aviation and reported separately (ipCC, 2006, p.3.78).

it is also recommended that a more detailed description of the methodology for splitting domestic and international fuel consumption be included in the next inventory report.


to improve road transport emission estimates, calculations should include the ability to compare emission estimates

3. energy Sector


using fuel consumption and kilometres travelled (based on travel data). this requires more knowledge of South Africa’sfleetprofile,andalsoanunderstandingofhowmuch fuel is consumed in the road transport sector as a whole. Furthermore, the development of local emission factors by fuel and vehicle-type will enhance the accuracy of the emission estimation.


National-level fuel consumption data are needed for estimating Co2 emissions for tier 1 and tier 2 approaches. in order to estimate CH4 and N2o emissions using a tier 2 approach, locomotives category-level data are needed. these approaches require that railway, locomotive companies or the relevant transport authorities provide fuel consumption data. the use of representative locally estimated data is likely to improve accuracy although uncertainties will remain large.


the provision of data by waterborne navigation is vital for the accurate estimation of emissions from this category. Complete and accurate data that will enable the consumption data to be split into domestic and international consumption, as well as the separate reporting of military consumption, would provide much improved emission estimates.

3.3.6 Other sectors [1A4]


3.3.6.1.1 Commercial/ institutional [1A4a]

this source category includes commercial/institutional buildings, as well as government, information technology, retail, tourism and services. there are great opportunities forimprovedenergyefficiencyinthebuildingsofthiscategory. this category consumes 14.8% of South africa’s

totalfinalenergydemand(DepartmentofEnergy,2008).Fuels included are residual fuel oil, other kerosene, gas/diesel oil, sub-bituminous coal, gas work gas and natural gas (Figure 3.10). liquid fuels contributed the most to the fuel consumption in this sector (75.5% in 2010) followed by solid fuels (24.3% in 2010).

3.3.6.1.2Residential[1A4b]

the residential sector includes fuel combustion in households. Fuels consumed in this category are other kerosene, residual fuel oil, lpG, sub-bituminous coal, wood/wood waste, other primary solid biomass and charcoal. in 2000 biomass fuel sources dominated (79.8%), however, from 2006 to 2010 there was no data reported for other primary solid biomass (Figure 3.11) therefore the biomass fuel source declined to 35.0% in 2010.

3.3.6.1.3Agriculture/forestry/fishing/fishfarms[1A4c]

the GHG emissions in this category include fuel combustion from agriculture (including large modern farms and small traditional subsistence farms), forestry, fishingandfishfarms.Fuelsincludedinthiscategoryaremotor gasoline, other kerosene, gas/diesel oil, residual fuel oil, lpG and sub-bituminous coal. liquid fuels dominate in this category (Figure 3.12). according to energy balance data (Department of energy, 2009), sub-bituminous coal was used only in 2000.



the estimation of total cumulative GHG emissions in the commercial/institutional category for the period 2000 to 2012 was 189 211 Gg Co2eq. emissions increased by 69.6 % over the 12 year period from 9 549 Gg Co2 eq in 2000 to 16 196 Gg Co2eq in 2012 (Figure 3.13). emissions were dominated by Co2 emissions, with a small percentage of CH4 and N2o.


Figure 3.11: Sector 1 energy: trend in fuel consumption in the residential category, 2000 – 2012.

Figure 3.10: Sector 1 energy: Fuel consumption in the commercial/institutional category, 2000 – 2012.

Natural Gas

Gas works gas

Sub-bituminous coal

residual Fuel oil

Gas / Diesel oil

other Kerosene

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

250 000

200 000

150 000

100 000

50 000

0

Charcoal

other primary solid biomass

Wood / Wood waste

Sub-bituminous coal

lpG

residual Fuel oil

other Kerosene

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

500 000

450 000

400 000

350 000

300 000

250 000

200 000

150 000

100 000

50 000

0

Fuel consumption (TJ)


3. energy Sector


Figure 3.13: Sector 1 energy: trend in total GHG emissions from other sectors, 2000 – 2012.

Figure3.12:Sector1Energy:Trendinfuelconsumptionintheagriculture/forestry/fishingcategory,2000–2012.

Sub-bituminous coal

lpG

residual Fuel oil

Gas / Diesel oil

other Kerosene

Motor Gasoline

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

50 000

45 000

40 000

35 000

30 000

25 000

20 000

15 000

10 000

5 000

0



in 2001 GHG emissions in this category increased by 15.6% compared with the previous year, and continued to increase annually from 2002 to 2003. that might have been as a result of economic growth and development during that period. the increase in 2004 (22.6%) was linked to the increased used of other kerosene. there was a decline in emissions in 2007 which was due to a reduction in sub-bituminous coal consumption that year. that was possibly linked to boiler-fuel switching from coal to gas in small-to-medium enterprises. GHG emissions in 2011 decreased by 9%, followed by an increase of 3.9% in 2012.


the estimation of total GHG emissions in the residential category for the period 2000 to 2012 was 211 817 Gg Co2eq. emissions in the residential sector have increased more than threefold from 7 101 Gg Co2eq in 2000 to 20 943 Gg Co2eq in 2012 (Figure 3.13). the increase was attributed mainly to population growth and economic growth (70.44% in 2000 to 75% in 2009). the GHG emissions in this category increased annually. the South african residential category consumes 20% of the total energy supply; this includes gas, electricity, candles, wood, dung,coal,LPG,paraffin,gasandothervegetablematter.in 2006, 72.8% of energy consumed by South african households was in the form of electricity, 29.1% in the form of coal, and 7.4% in the form of petroleum products (suchasLPGandparaffin)(DepartmentofEnergy,2009b).By 2009, 75% of households (9 245 357 households) were electrified(DepartmentofEnergy,2009b).

Ithasbeenrecordedthatmorethan10millionelectrifiedhouseholdsinSouthAfricahavebeenfittedwitheightincandescent lights per household. in 2008 eskom rolled outcompactfluorescentlamps(CFLs),whichresultedina saving of 800 MW of electricity. By September 2009, more than 30 million lamps had been replaced. energy consumed by households represented 17% of the country’s net use. Most of household energy was obtained from fuel wood (50% of the net household energy), primarily in the

rural areas, with the remainder being obtained from coal (18%),illuminatingparaffin(7%)andasmallamountfromlpG. the estimated number of households with access to electricity increased from 4.5 million (50.9%) in 1994 to 9.1 million (73%) in 2008. Coal is used by approximately 950 000 households countrywide.

GHG emissions in this category increased annually, however in 2011 there was a decline of 16.8%. this could have been due to an increase of more than 10% in the food price. the price of basic foodstuffs, such as maize, wheat, soya beans and rice, increased as a result of changing climatic conditions and rising demand. in 2007 GHG emissions in the residential category increased by 13.6% compared with the previous year.


primary agriculture contributes approximately 3.2% to the GDp of South africa and provides almost 9% of formal employment. the majority of the energy for agriculture is sourced from diesel and vegetable waste (Department of energy, 2010). in 2006 approximately 69% of energy for use in agriculture was sourced from petroleum products, 29.9% from electricity and 1.1% from coal (Department of energy, 2010). the estimated total GHG emissions oftheagriculture/forestry/fishing/fishfarmscategoryforthe period 2000 to 2012 was 36 801 Gg Co2eq. GHG emissions increased from 2 387 Gg Co2eq in 2000 to 3 432 Gg Co2 eq in 2012, with annual increases of up to 9.3%. Decreases occurred in 2001 and 2008 (Figure 3.13). GHG emissions in 2008, during the global economic crisis, decreased by 1.7%, while in 2010 GHG emissions peaked, increasing by 7.9%.

there was a decline in emissions between 2000 and 2001. Thecontributionoftheagriculture/forestry/fishing/fishfarms category to the GDp also decreased by 3.3% during the same period. in 2008 the sector’s GHG emissions decreased by 1.7%, accompanied by a massive GDp contribution of 16.1% in the same period. However, in 2009 its contribution to the GDp decreased to 0.3% and

3. energy Sector


the GHG emissions increased by 1.46%., mainly because of the global economic crisis in 2008/09.


the tier 1 approach was used to estimate emissions from all the other sectors. to estimate the total GHG emissions of this sector, the amount of fuel combusted was multiplied with the default emission factors from the 2006 ipCC Guidelines (Table 3.17).



Data on fuel consumption in the commercial/institutional buildings category was sourced from the Doe’s energy digest reports, the DMr’s SaMi report (solid fuels and natural gas) and Sapia (liquid fuels). the Doe energy reports were used to source solid fuels for the period 2000 to 2006, for the period 20007 to 2010, the SaMi report was used to extrapolate the consumption of solid fuels for this category. an NCv of 0.0243 tJ/tonne was used to convert fuel quantities into energy units (Department of energy, 2009a).


Data on fuel consumption in the residential sector was obtained from the Doe’s energy digest reports, the DMr’s SaMi report, the Fao and Sapia. the Doe energy reports were used to source solid fuels for the period 2000 to 2006, for the period 2007 to 2010), the SaMi report was used to extrapolate the consumption of solid fuels for this category. an NCv of 0.0243 tJ/tonne was used to convert fuel quantities into energy units (Department of energy, 2009a).


Data on fuel consumption in the agriculture, forestry, fishingandfishfarmscategorywasobtainedfromtheDoe’s energy digest (mostly solid fuels) and Sapia for liquid fuels. the trend for the consumption of fuels in this category has been increasing and decreasing throughout the period 2000 to 2010. according to the Doe’s energy balances, solid fuels in the form of sub-bituminous coal were only consumed in 2000.


a country specific emission factor for Co2 for sub-bituminous coal was applied (see section 3.3.3.4). this countryspecificEFwasusedforallyearsgoingbackto2000. For all other fuels the ipCC 2006 Guideline default emission factors were used to determine emissions from all other sectors (Table 3.17).

table 3.17: Sector 1 energy: emission factors used for all other sectors (Source: 2006 ipCC Guidelines).


Co2 CH4 N20





lpG 63 100 1 0.1


Gaswork gas 44 400 1 0.1

Natural gas 56 100 1 0.1

Wood/wood waste 112 000 30 4

other primary solid biomass

100 000 30 4

Charcoal 112 000 30 4



the uncertainties in Co2 emissions are relatively low in fossil fuel combustion. these emission factors are determined by the carbon content of the fuel. emission factors for CH4andmorespecificallyN2o are highly uncertain. the uncertainty on the CH4 emission factor is 50 to 150%, while for N2o it is an order of magnitude. this high uncertainty is due to the lack of relevant and accuratemeasurementsand/orinsufficientunderstandingof the emission generating process.


an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and to identify areas of improvements.


in the previous 2000 GHG inventory, the activity data for this category was sourced from the energy balances published by the Doe. For this inventory a combination of sources of activity data were used, including the Doe’s energy digest, the DMr’s SaMi report, Sapia and the Fao. the inclusion of various sources of data has improved the accuracy of the GHG inventory which led to a recalculation of emissions for these categories. Further, the N2o emission factor for coal was corrected. in the previous inventory a value of 3 kg tJ-1 was applied but the source of the data could not be traced so the default ipCC emission factor was applied. the results for the published 2000 GHG inventory had to be recalculated as a result of the availability of more robust activity data.


there are several opportunities for improvement in this category including the collection of additional activity data,identificationanddisaggregationofcontributing

sources in each section, and the development of source specificmethodologies.


the tier 1 approach is used for the simplest calculation methods or methods that require the least data; therefore this approach provides the least accurate estimates of emissions. the tier 2 and tier 3 approaches require more detailed data and resources to produce accurate estimates of emissions. the commercial/institutional sector should be willing to co-operate in the provision of data for the purposes of inventories. a regulatory framework should be established and implemented to ensure that sectors provide the data necessary for the compilation of the inventory.


investigations and studies of the residential sector in South africa are necessary for the accurate estimation of emissions. Due to the many households, uniform reporting would be possible if data were collected by local government.


a regulatory framework should be established and implemented to ensure that sectors provide the data necessary for the compilation of the inventory.

3.3.7 Non-specified [1A5]


this category refers to all remaining emissions from fuel combustionthatwerenotspecifiedelsewhereinthisdocument. it should include emissions from fuel delivered to the military in the country and delivered to the military of other countries that are not engaged in multilateral operations.

3. energy Sector


3.3.7.1.1 Stationary [1A5a]

this section includes emissions from fuel combustion instationarysourcesthatarenotspecifiedelsewhere.the only fuel reported under this category was the consumption of motor gasoline.


Thenon-specifiedstationarycategoryshowedasteadyincrease of 12.7% in total GHG emissions between 2000 and 2012. emissions were estimated at 989 Gg Co2eq in 2000 and 1 114 Gg Co2eq in 2012 (Table 3.18). there was a slight decline of 0.71% between 2000 and 2002 and a 4.23% decline in 2008.

Table3.18:Sector1Energy:TrendinconsumptionandGHGemissionsfromthenon-specifiedsector,2000–2012.

PeriodConsumption CO2 CH4 N2O Total GHG

(tJ) (Gg) (Gg Co2eq) (Gg Co2eq) (Gg Co2eq)

2000 14 222 986 1.07 2.54 989

2001 14 145 980 1.06 2.53 984

2002 14 138 980 1.06 2.53 983

2003 14 592 1 011 1.01 2.59 1 015

2004 15 027 1 041 1.13 2.69 1 045

2005 15 274 1 058 1.15 2.73 1 062

2006 15 430 1 069 1.16 2.76 1 073

2007 15 811 1 096 1.19 2.83 1 100

2008 15 142 1 049 1.14 2.71 1 053

2009 15 476 1 072 1.16 2.77 1 076

2010 16 376 1 135 1.23 2.93 1 139

2011 16 365 1 134 1.23 2.93 1 138

2012 16 025 1 111 1.2 2.87 1 114

3.3.7.3 Methodologica Issues

the tier 1 approach was used for the calculation of emissionsinthenon-specifiedsector.Toestimatethetotal GHG emissions for this sector, the activity data (fuel consumed) was multiplied by the default emission factor from the 2006 ipCC Guidelines.


Dataonfuelconsumptioninthenon-specifiedcategorywere sourced from the Doe’s energy digest reports (solid fuels and natural gas), the SaMi report to extrapolate activity data for solid fuels for the period 2007 to 2012, and Sapia (liquid fuels). the NCvs applied for the conversion of fuel quantities into energy units were sourced from the digest of energy statistics report (Department of energy, 2009a).



ipCC default emission factors from the 2006 ipCC Guidelines were used to estimate GHG emissions from thenon-specifiedsector(Table 3.19).


the uncertainties in Co2 emissions are relatively low in fossil fuel combustion. these emission factors are determined by the carbon content of the fuel. emission factors for CH4and,morespecifically,N2o are highly uncertain. this high uncertainty is due to the lack of relevantandaccuratemeasurementsand/oraninsufficientunderstanding of the emission-generating process.


an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvements.


Inthe2000GHGinventory,thenon-specifiedcategorywas excluded from the estimations. For this inventory the Doe’s energy digest and Sapia were used as sources of data for this category to ensure completeness by including all categories as recommended by the ipCC Guidelines. the results for the published 2000 GHG inventory had to be recalculated as a result of availability of more robust activity data.

table 3.19: Sector 1 energy: emission factors for calculating emissions fromthenon-specifiedsector(Source:2006IPCCGuidelines).


Co2 CH4 N20


3.3.7.8 Source-specific Planned Improvements and Recommendations

the tier 1 approach is used for the simplest calculation methods or methods that require the least data; therefore this approach provides the least accurate estimates of emissions. the tier 2 and tier 3 approaches require more detailed data and resources to produce accurate estimates of emissions. Sourcing of activity data for pipeline transport, and fuel consumption associated with ground activities at airports and harbours is planned for the next inventory compilation cycle.

3.4 Fugitive emissions from fuels [1b]

Fugitive emissions refer to the intentional and unintentional release of GHGs that occur during the extraction, processing and delivery of fossil fuels to the pointoffinaluse.Methaneisthemostimportantemissionsourced from solid fuels fugitive emissions.

in coal mining activities, the fugitive emissions considered were from the following sources:

• Coal mining, including both surface and underground mining;

• Coal processing;

• the storage of coal and wastes; and

• the processing of solid fuels (mostly coal)

3.4.1 Solid fuels [1b1]


3.4.1.1.1 1B1a Coal mining and handling [1B1a]

the geological processes of coal formation produce

3. energy Sector


CH4 and Co2. CH4 is the major GHG emitted from coal mining and handling. in underground mines, ventilation causessignificantamountsofmethanetobepumpedintothe atmosphere. Such ventilation is the main source of CH4 emissions in hard coal mining activities. However, methane releases from surface coal mining operations are low. in addition, methane can continue to be emitted from abandoned coal mines after mining has ceased.

according to the 2006 ipCC Guidelines, the major sources for the emission of GHGs for both surface and underground coal mines are:

• Mining emissions: the release of gas during the breakage of coal and the surrounding strata during mining operations

• Post-mining emissions: emissions released during the handling, processing and transportation of coal. Coal continues to emit gas even after it has been mined, but at a much slower rate than during coal breakage stage.

• Low-temperature oxidation: emissions are released when coal is exposed to oxygen in air; the coal oxidizes to slowly produce Co2.

• Uncontrol led combust ion: Uncontrol led combustion occurs when heat produced by low-temperature oxidation is trapped. this type of combustion is characterized by rapid reactions, sometimesvisibleflamesandrapidCO2 formation. it may be anthropogenic or occur naturally.



in 2000 South african mines produced 217.5 Mt of coal and the fugitive emissions in that period were 2 003 Gg Co2eq. in 2004, 242.82 Mt of coal were produced, of which 146.27 Mt were consumed locally and the fugitive emissions were equivalent to 2 141 Gg Co2eq. in 2005,

245 Mt of coal were produced, and 174 Mt was consumed locally. in 2006, 246 Mt were produced and the fugitive emissions accounted for 2 154 Gg Co2eq. in 2007, 247.7 Mt of coal were produced (with a value of r14.69 billion) and 2 179 Gg Co2eq of fugitive emissions were produced.

total GHG fugitive emissions from coal mining decreased from 2 003 Gg Co2eq in 2000 to 1 760 Gg Co2eq in 2012 (Table 3.20). GHG emissions in this category increased by 49.1% between 2000 and 2011, then declined in 2012. the increase is mainly due to the increased demand for coal, particularly for electricity generation. Since opencast mining dominates overall coal production, CH4

emissions have remained relatively stable over the 2000 to2012timeseries.Country-specificemissionfactorshaveconfirmedthatSouthAfricancoalseamshavelittletrapped CH4 in situ.

table 3.20: Sector 1 energy: Fugitive emissions from coal mining for the period 2000 to 2012.

PeriodCO2 CH4 Total GHG

Gg (Gg Co2eq) (Gg Co2eq)

2000 24 1 979 2 003

2001 23 2 137 2 161

2002 23 1 938 1 961

2003 25 2 093 2 118

2004 25 2 141 2 167

2005 26 2 155 2 181

2006 26 2 154 2 180

2007 26 2 179 2 205

2008 26 2 219 2 245

2009 26 2 204 2 230

2010 27 2 239 2 266

2011 35 2 951 2 986

2012 21 1 739 1 760




a tier 2 approach was used to calculate fugitive emissions from coal mining and handling. Fugitive emission estimates were based on coal production data. Coal waste dumps were also considered as an emission source. this methodology requires coal production statistics by mining type (above-ground and below-ground) and this split (53% surface mining and 47% underground mining) was based on the SaMi report for 2008. it was assumed that the split remained constant throughout the entire time series.



Data on coal production (Table 3.21) was obtained from the SaMi report compiled by the DMr (SaMi, 2009) and Coaltech.


Countryspecificemissionfactorsweresourcedfromthestudy done by the local coal research institute (DMe, 2002). this study has showed that emission factors for the SouthAfricancoalminingindustryaresignificantlylowerthan the ipCC default emission factors (Table 3.22).

the 2006 ipCC Guidelines do not provide Co2 emission factors related to low-temperature oxidation of coal, however,SouthAfricahasdevelopedcountry-specificCo2 emission factors for this and, therefore, has estimated emissions related to this activity.


the major source of uncertainty in this category is activity data on coal production statistics. according to the 2006 IPCCGuidelines,country-specifictonnagesarelikelytohave an uncertainty in the 1 to 2% range, but if raw coal data are not available, then the uncertainty will increase to about ±5%, when converting from saleable coal

table 3.21: Sector 1 energy: Coal mining activity data for the period 2000 to 2012.

PeriodOpencast Underground

(tonnes)

2000 152 430 357 135 174 090

2001 151 473 376 134 325 446

2002 149 287 553 132 387 075

2003 161 217 666 142 966 609

2004 164 944 899 146 271 891

2005 166 040 627 147 243 575

2006 165 935 025 147 149 928

2007 167 855 716 148 853 182

2008 170 937 442 151 586 034

2009 169 791 125 150 569 488

2010 172 502 123 152 973 581

2011 153 430 367 113 400 000

2012 152 430 368 118 800 000

3. energy Sector


productiondata.Thedataarealsoinfluencedbymoisturecontent, which is usually present at levels between 5 and 10 %, and may not be determined with great accuracy.


an inventory compilation manual documenting sources of data, data preparation and sources of emission factors was used to compile emission estimates for this source category.Emissionestimateswerealsoverifiedwithemission estimates produced by the coal mining industry. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvements.


emissions were recalculated for 2000 using country-specificEFs.Therecalculationresultedinthereductionof emissions from 40 Mt Co2eq to 2 Mt Co2eq, resulting in a 95% reduction in the emission estimate.


More attention needs to be placed on the collection of fugitive emissions from abandoned mines and the spontaneous combustion of underground coal seams.

3.4.2 Oil and natural gas [1b2]


the sources of fugitive emissions from oil and natural gas included, but were not limited to, equipment leaks,evaporationandflashinglosses,venting,flaring,incineration and accidental losses (e.g. tank, seal, well blow-outs and spills) as well as transformation of natural gas into petroleum products.


the total estimation of cumulative GHG emissions from

Table3.22:Sector1Energy:Comparisonofcountry-specificandIPCC2006defaultemissionfactorsforcoalmining.

Mining method Activity GHGEmission factor (m3 tonne-1)

SouthAfricaspecific 2006 ipCC default

Underground Mining

Coal Mining

CH4

0.77 18

post-mining (handling and transport)

0.18 2.5

Surface Mining

Coal mining 0 1.2

post-mining (storage and transport)

0 0.1

Underground Mining

Coal mining

Co2

0.077 Na


0.018 Na

Surface Mining

Coal mining 0 Na


0 Na


flaringwasequivalentto13496GgCO2 for the period 2000 to 2012. emissions decreased by 20% from 752 Gg Co2eq in 2000 to 642 Gg Co2eq in 2012 (Table 3.23).


Fugitive emissions are a direct source of GHGs due to the release of CH4 and formation Co2 (Co2 produced in oil and gas when it leaves the reservoir). Use of facility-level production data and facility-level gas composition and vent flowrateshasfacilitatedtheuseofTier3methodology.Hence, Co2emissionsfromventingandflaringhavebeenestimated using real continuous monitoring results and therefore no emission factors were used.


Thisisthefirsttimethissubcategoryhasbeenaccountedfor in the national GHG inventory. Data on oil and natural gasemissionswereobtaineddirectly fromrefineriesand, to a lesser extent, from the energy digest reports (Department of energy, 2009a).


according to the 2006 ipCC Guidelines, gas compositions are usually accurate to within ±5% on individual components. Flow rates typically have errors of ±3% or less for sales volumes and ±15% or more for other volumes. Given that the activity data used is sourced at facility level, the uncertainty is expected to be less than 3%.




Thisisthefirsttimethatthiscategoryhasbeenincludedso no recalculations were necessary.


to improve the completeness of the accounting of emissions from this subsector, future activity data collection activities need to focus on upstream natural gas production and downstream transportation and distribution of gaseous products.

3.4.3 Other emissions from energy production [1b3]


according to the 2006 ipCC Guidelines (p.4.35), other emissions from energy production refers to emissions from geothermal energy production and other energy production not included in the 1.B.1 and/or 1.B.2 categories. in the South african context, this refers to the coal-to-liquid (Ctl) and gas-to-liquid (Gtl) processes.

Table3.23: Sector1Energy:TotalGHGemissionsfromventingandflaringfortheperiod2000–2012.

Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Gg Co2eq

752 753 955 1 458 1 379 1 160 1 133 1 133 1 138 1 243 964 786 642

3. energy Sector


these GHG emissions are most specifically fugitive emissions related to the two mentioned processes (Ctl and Gtl) with the emphasis on Co2 removal.


the total estimation of cumulative GHG emissions from other emissions from energy production is equivalent to 374 300 Gg Co2eq for the period 2000 to 2012. emissions fluctuatedthroughoutthe12yearperiod(Table 3.24).


the use of facility-level production data and facility-level gascompositionandventflowratesenabledtheuseoftier 3 methodology. Hence, Co2 emissions from other emissions from energy production have been estimated using real continuous monitoring results and material balances.


Data on other emissions from energy production were obtained from both Sasol and petroSa.


No source-specif ic uncertainty analysis has been performed for this source category. Currently, uncertainty data does not form part of the data collection and measurement programme. this is an area that will

require improvement in future inventories. Facilities are to be encouraged to collect uncertainty data as part of data collection and measurement programmes. time-series activity data was validated using information on mitigation projects that have been implemented in the past 10 years and other factors such as economic growth and fuel supply and demand.


Quality assurance is currently done in an ad-hoc manner: the department reviews the material balance and measurement data supplied by facilities. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures which were followed for the compilation of this inventory and identify areas of improvement.


Nosource-specificrecalculationsweredoneforthissection.


No improvements are planned for this section.

table 3.24: Sector 1 energy: total GHG emissions from the category other emissions from energy production (1B3), 2000 – 2012.

Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Gg Co2eq

30 597 31 039 31 317 29 812 31 568 27 854 27 708 28 160 27 134 27 435 27 019 26 836 27 821


4.1 An overview of the IPPU sector

the ippU sector includes GHG emissions sourced from industrial processes, the use of GHG emissions in products and the use of fossil fuels (non-energy uses). the main emission sources are releases from industrial processes that chemically or physically transform raw material (e.g. ammonia products manufactured from fossil fuels). GHG emissions released during these processes are Co2, CH4, N2o, HFCs, SF6 and pFCs. also included in the ippU sector are GHG emissions used in products such as refrigerators, foams and aerosol cans.

the estimation of GHG emissions from non-energy sourcesisoftendifficultbecausetheyarewidespreadanddiverse.ThedifficultiesintheallocationofGHGemissionsbetween fuel combustion and industrial processes arise when by-product fuels or waste gases are transferred from the manufacturing site and combusted elsewhere in different activities. the largest source of emissions in the ippU sector in South africa is the production of iron and steel.


Major GHGs generated by the ippU sector include Co2, N2o, CH4 and pFCs. the main emissions sources for this category are as follows:

• Manufacture of mineral products, mainly cement;

• Manufacture of chemical products, such as nitric acid and adipic acid; and

• Metal production, mainly iron and steel.

the performance of the economy is the key driver for trends in the ippU sector. the South african economy is directly related to the global economy, mainly through exportsandimports.SouthAfricaofficiallyenteredan

economicrecessioninMay2009,whichwasthefirstin17years. Until the global economic recession affected South africa in late 2008, economic growth had been stable and consistent. according to Statistics South africa, the GDp increased annually by 2.7%, 3.7%, 3.1%, 4.9%, 5.0%, 5.4%, 5.1% and 3.1% between 2001 and 2008, respectively. However in the third and fourth quarters of 2008, the economy experienced enormous recession, and this continuedintothefirstandsecondquartersof2009.as a result of the recession, GHG emissions during that period decreased enormously across almost all categories in the ippU sector.

in 1990 ippU GHG emissions accounted for 8.9% of South africa’s total GHG emissions (excl. FolU), whereas in 1994 and 2000 it contributed 8.0% and 6.5%, respectively. When analysing the recalculated ippU sector emissions for 2000, there was an increase of 4.6% when compared to 1990. Between 2000 and 2010 there was a 1.2% decrease in overall emissions from the ippU sector, although there were increases during this period.

TheGHGemissionsintheIPPUsectorfluctuatedduringthe 12-year reporting period (Figure 4.1). accumulated emissions totalled 467 127 Gg Co2eq. ippU sector emissions increased annually by an average of 2.7% between 2000 and 2006 due to robust economic growth during this time, which led to an increased demand for products. Between 2006 and 2009 there was a decline of 16% (from 39 298 Gg Co2eq to 37 128 Gg Co2eq) in the ippU emissions. this decrease was mainly due to the global economic recession and the electricity crisis that occurred in South africa during that period. in 2010 emissions increased again by 7.4%. the economy was beginning to recover from the global recession. another reason for the increase in GHG emissions in 2010 was that South africa hosted the 2010 FiFa World Cup and, as a result, an increase in demand for commodities was experienced.

4. INDUSTRIAL PROCESSES AND OTHER PRODUCT USE

4. industrial processes and other product Use


Themostsignificantsourceofemissions intheIPPUsector was the metal industries subsector, which contributed between 79.9% and 80.5% over the period 2000 to 2012. the biggest contributor in 2012 to the metal industry emissions was the iron and steel industry (50.5%). Co2 emissions constituted between 92.2% and 88.7% of the total ippU emissions between 2000 and 2012, while HFCs and pFCs contributed an average of 2.2% and 2.7% respectively (Figure 4.2).

4.1.2 Key sources

the major key category in the ippU sector in terms of

emissions is iron and steel production (2.C.1). the other key categories are ferroalloys production (2.C.2), product usesassubstitutesforODS (2.F.1), aluminium production (2.C.3) and chemical industries (2.B) (see Table 1.9 and Table 1.11).Thesearesimilarfindingstothekeycategoryanalysis performed in the 2010 inventory.

4.1.3 Completeness

the ipCC Guidelines recommend that the national GHG inventory should include all relevant categories of sources, sinks and gases. the completeness of inventories refers to completeness of all gases, completeness of sources and

Figure 4.1: Sector 2 ippU: trend and emission levels of source categories, 2000 – 2012.


Figure 4.2: Sector 2 ippU: trend and emission levels of the various greenhouse gases, 2000 – 2012.



sinks categories, completeness of geographical coverage and completeness in the coverage of the years in the time series.

in the compilation of the GHG emissions inventory it is important to minimize omissions and also avoid double counting emissions. therefore it is imperative to ensure thatallthesourcesthathavebeenidentifiedareallocatedto the appropriate sources. in the compilation of GHG emissions from the ippU sector, it is important that all significantGHGemissionsfromnon-energyusesoffossilfuels are reported, without any double counting. the sum of these emissions includes (a) fuels used as feedstock in the chemical industry, (b) fuels used as reductants in the metal industry, (c) fuel products oxidized during use (partly or fully; direct emissions or emissions of carbon-containing non-Co2 gases – NMvoC, Co and CH4 – oxidized in the atmosphere).

in the completion of this inventory the main challenge was lack of activity data or costs associated with gathering the activity data. it is good practice to include all the sources of GHG emissions in a country, if actual emission quantities have not been estimated or are not reported then it should be transparent in the inventory report. this inventory is not complete as it does not include all categories and gases listed in the ipCC Guidelines. the reasons for not including some categories are provided in Table 4.1. identifying the categories that have not been included in the inventory will give direction on the disproportionate amount of effort required for the collection of data in this sector. the structure is based on the naming and coding of the 2006 ipCC Guidelines and the Common reporting Format (CrF) used by the UNFCCC.

Table4.1: Sector2IPPU:Classificationofcategoriesofemissionsexcludedfromthisinventory.

Category Definition of category Justification for exclusion

2A Mineral industry 2a4 other process uses of carbonatesNot estimated (Ne): emissions occur but have not been estimated or reported because of lack of data.

2b Chemical industry

2B3 adipic acid productionNot occurring (No): an activity or process does not exist within a country.

2B4 Caprolactam, glyoxal and glyoxylic acid production

Not occurring (No): an activity or process does not exist within a country.

2B7 Soda ash productionNot occurring (No): an activity or process does not exist within a country.

2B9 Fluorochemical productionNot occurring (No): an activity or process does not exist within a country.

2C Metal industry 2C4 Magnesium productionNot occurring (No): an activity or process does not exist within a country.


Category Definition of category Justification for exclusion

2E Electronics

2e1 integrated circuit or semi-conductor

Not estimated (Ne): emissions and/or removals occur but have not been estimated or reported.

2E2TFTflatpaneldisplay

2e3 photovoltaics

2E4Heattransferfluid

2G Other product manufacture and use

2G1 electrical equipment2G2 SF6 and pFCs from other product uses 2G3 N2o

Not estimated (Ne): emissions and/or removals occur but have not been estimated or reported.

2H Other2H1 pulp and paper industry Not estimated (Ne): emissions and/

or removals occur but have not been estimated or reported.2H2 Food and Beverages industry

4.2 Mineral Production [2A]

4.2.1 Source category description

Mineral production emissions are mainly process-related GHG emissions resulting from the use of carbonate raw materials. the mineral production category is divided into fivesubcategories;cementproduction,limeproduction,glass production, process uses of carbonates, and other mineral products processes. For this inventory report, emissions are reported for three subcategories: cement production, lime production and glass production.

4.2.1.1 Cement Production [2A1]

the South african cement industry’s plants vary widely in age,rangingfromfivetoover70years(DMR,2009).Themost common materials used for cement production are limestone, shells, and chalk or marl combined with shale, clay, slate or blast-furnace slag, silica sand, iron ore and gypsum. For certain cement plants, low-grade limestone appears to be the only raw material feedstock for clinker production (DMr, 2009). portland cement, which has a clinker content of >95%, is described by the class CeM i. CeM ii cements can be grouped depending on their clinker

content into categories a (80 – 94%) and B (65 – 79%). portland cement contains other puzzolanic components suchasblast-furnaceslag,microsilica,flyashandgroundlimestone. CeM iii cements have a lower clinker content and are also split into subgroups: a (35 – 64% clinker) and B (20 – 34% clinker). South africa’s cement production plants produce portland cement and blended cement products, such as CeM i, and, more recently, CeM ii and CeM iii. Cement produced in South africa is sold locally and to other countries in the Southern africa region, such as Namibia, Botswana, lesotho and Swaziland.

the main GHG emission in cement production is Co2 emitted through the production of clinker, an intermediate stage in the cement production process. Non-carbonate materials may also be used in cement production, which reduce the amount of Co2 emitted. However, the amounts of non-carbonate materials used are generally very small and not reported in cement production processes in South africa. an example of non-carbonate materials would be impurities in primary limestone raw materials. it is estimated that 50% of the cement produced goes to the residential building market (DMr, 2009); therefore, any changes in the interest rates that affect the residential market will affect cement sales.



4.2.1.2 Lime Production [2A2]

lime is the most widely used chemical alkali in the world. Calcium oxide (Cao or quicklime or slacked lime) is sourced from calcium carbonate (CaCo3), which occurs naturally as limestone (CaCo3) or dolomite (MgCo3). Cao is formed by heating limestone at high temperatures to decompose the carbonates (ipCC, 2006, 2.19) and produce Cao. this calcination reaction produces Co2 emissions. lime kilns are typically rotary-type kilns, which are long, cylindrical, slightly inclined and lined with refractory material. at some facilities, the lime may be subsequently reacted (slaked) with water to produce hydrated lime.

in South africa the market for lime is divided into pyrometallurgical and chemical components. Hydrated lime is divided into three sectors: chemical, water purificationandothersectors(DMR,2010).Quicklimeand hydrated lime contributed an average of 92% and 8%, respectively (DMr report r85/2010). lime has wide applications, e.g., it is used as a neutralizing and coagulating agent in chemical, hydrometallurgical and water treatment processes and a fluxing agent in pyrometallurgical processes. pyrometallurgical quicklime sales have been increasing, while the demand for quicklime in the chemical industry has been decreasing (DMr, 2010).

4.2.1.3 Glass Production [2A3]

there are many types of glass and compositions used commercially, however the glass industry is divided into fourcategories:containers,flat(window)glass,fibreglassand speciality glass. When other materials (including metal) solidify, they become crystalline, whereas glass (a super cool liquid) is non-crystalline. the raw materials used in glass production are sand, limestone, soda ash, dolomite, feldspar and saltcake. the major glass raw materials which emit Co2 during the melting process are limestone (CaCo3), dolomite CaMg (Co3)2 and soda ash (Na2Co3). Glass makers do not produce glass only from raw materials, they also use a certain amount of

recycled scrap glass (cullet). the chemical composition of glass is silica (72%), iron oxide (0.075%), alumina (0.75%), magnesium oxide (2.5%), sodium oxide (14.5%), potassium oxide (0.5%), sulphur trioxide (0.25%) and calcium oxide (7.5%) (pFG glass, 2010).


the cumulative GHG emissions from the mineral industry for the reporting period 2000 to 2012 were 59 472 Gg Co2eq. emissions increased by 35% between 2000 and 2007 to 5 217 Gg Co2eq, after which emissions slowly declined (by 15.8%) to 4 411 Gg Co2eq in 2012 (Figure 4.3). over the 12-year period the emissions from the mineralindustriesincreasedby14.7%.ThemostsignificantGHG emission sources in 2012 were cement production (contributing 87.2% of total GHG emissions), followed by lime production (10.3%). GHG emissions from glass production accounted for only 2.6% of the total emissions.


the GHG emissions from cement production increased from 3 347 Gg Co2eq in 2000 to 4 583 Gg Co2eq in 2007, after which emissions declined to 3 845 Gg Co2eq in 2012. CementproductioninSouthAfricaincreasedsignificantlyfrom the period 2000 to 2007 as a result of economic growth. there was a 2.0%, 7.1%, 5.8% and 2.5% decrease in emissions in 2008, 2010, 2011 and 2012, respectively. in 2009 the South african economy went into recession and the GDp decreased by 1.8% in that year. Cement demand in the residential market and construction industry in 2009/2010 decreased due to higher interest rates, increased inflationandthe introductionoftheNational Credit act (DMr, 2009/2010). one reason for the decrease in GHG emissions in 2008 is that projects with an estimated value of r4 billion and r6 billion were postponed or cancelled until September 2008 as a result of electricity supply constraints (association of Cement and Concrete institute, 2008). Demand for cement since 2009 has continued to increase. in 2012, the DMr


estimated that cement imports increased by 62% in 2012 compared with 2011 imports (DMr 2013). Hence, despite an increase in cement demand, emissions have declined by 15% between 2009 and 2012 because clinker is being imported.


the demand for lime production in South africa is mainly linked to developments and investments in the steel and metallurgical industries (DMr report r85/2010). the local sales volume of lime decreased by 21% in 2012, compared with the previous year, to 1.2 Mt due to a decrease in demand for quicklime for the steel and alloys industries. according to media reports, South african steelmakers collectively produced 6.9 Mt in 2012 compared with 7.5 Mt the previous year, dropping in the world steel-production

rankings from 21 to 22. the total sales volume of quicklime for pyrometallurgical and chemical applications decreased by 22% to 1.11 Mt owing to subdued demand from the steel industry, and sales value decreased by 14% to r1.1 billion (DMr 2013). the total cumulative estimation of GHG emissions from lime production for the period 2000 to 2012 was 6 480 Gg Co2eq.

Thefluctuationsinlimeproductionweredirectlylinkedto developments and investments in the steel and metallurgical industries. in 2009 the lime industry declined as a result of the economic recession, hence, in that year, GHG emissions from lime production decreased by 3.4%. overall, the GHG emissions from this category increased by 6.3% between 2000 and 2012, which is mainly due to an increase in the number of infrastructure projects and the 2010 FiFa World Cup preparation.

Figure 4.3: Sector 2 ippU: trend and emission levels in the mineral industries, 2000 – 2012.




South africa’s glass production emissions increased consistently from 2001 to 2005 and again in 2007 to 2008. in 2009 and 2010 GHG emissions from glass production declined by 6.6% and 5.3%, respectively. this decline in emissions was mainly because of the global economic crisis which affected the glass-manufacturing market. GHG emissions picked up again in 2011 (1.9%) and 2012 (7.5%). over the 12-year period there was a 53.2% increase in emissions from glass production. the glass-manufacturing marketislargelyinfluencedbyconsumerbehaviourandconsumer spending; therefore any negative changes in the economy will affect the glass manufacturing industry. the total cumulative estimation of GHG emissions from glass production for the period 2000 to 2012 was 1 295 Gg Co2eq.

4.2.3 Methodological issues


a tier 2 approach, based on plant-level data, was used to determine the emissions from cement production.


the production of lime involves various steps, which include the quarrying of raw materials, crushing and sizing, calcining the raw materials to produce lime, and (if required) hydrating the lime to calcium hydroxide. the tier 1 approach was used for the calculation of GHG emissions from lime production. this report estimated the total lime production based on the aggregate national value of the quantity of limestone produced, using the breakdown of the types of lime published in the SaMi report (2012/2013). Based on the ipCC’s default method, an emission factor that assumes an 85% to 15% ratio of limestone to dolomite was used.


the tier 1 approach was used to determine estimates of the GHG emissions from glass production. the default ipCC emission factor was used and the cullet ratio for national level glass production was also determined from industry supplied activity data.

4.2.4 Data sources


Data on cement production in South africa was obtained from the cement production industry (Table 4.2). according to the association of Cementitious Material producers (2008), cement demand declined in South africa between 2008 and 2012. With the decline in domestic demand of cement due to economic recession and the electricity crisis, the cement production industry had to reduce production. local cement production declined by 14 % between 2008 and 2012. in 2009 Sephaku Cement entered the local cement production market, adding an estimated 1.2 Mt of production capacity to the market (SaMi, 2012/2013).


For the calculation of GHG emissions in cement production, Co2 emission factors were sourced from the 2006 ipCC Guidelines. it was assumed that the Cao composition (one tonne of clinker) contains 0.65 tonnes of Cao from CaCo3. this carbonate is 56.03% of Cao and 43.97% of Co2 by weight (ipCC, 2006, p. 2.11). the emission factor for Co2, provided by ipCC 2006 Guidelines, is 0.52 tonnes of Co2 per tonne clinker. the ipCC default emission factors were used to estimate thetotalemissions.Thecountry-specificclinkerfractionfor the period 2000 to 2012 ranged between 69% - 76% (Table 4.3).



the DMr publishes data on limestone and dolomite production in South africa in the SaMi report (DMr, 2012/2013). the SaMi provides a breakdown of limestone demand; 80% goes to cement manufacturing and the remaining 20% goes to metallurgical, agricultural and other uses. No data was provided for lime production, therefore it was assumed that the ‘other’ (6.0%) in the breakdown of limestone demand goes to lime production.


For the calculation of GHG emissions from lime production, GHG emission factors were sourced from the 2006 ipCC Guidelines. in South africa data was acquired for high-calcium lime, which has a range of 93 to 98% Cao content, and hydraulic lime, which has a range of 62 to 92% Cao. the GHG emission factor for high-calcium lime (0.75 tonnes Co2/tonne Cao) and hydraulic lime (0.59 tonnes Co2 per tonne Cao) were used for quicklime

table 4.2: Sector 2 ippU: activity data for cement, lime and glass production, 2000 – 2012.

Period

Cement Production

Quicklime Hydrated limeTotal Glass Produced

Recycled Glass

tonnes tonnes tonnes tonnes tonnes

2000 6 436 640 532 100 46 270 561 754 189 957

2001 6 551 695 522 910 45 470 624 156 202 227

2002 6 451 258 572 369 49 771 667 110 225 382

2003 6 879 106 586 969 51 041 702 008 245 359

2004 7 404 575 608 056 52 874 726 644 247 182

2005 8 054 255 685 860 59 640 775 839 264 026

2006 8 627 144 755 302 65 678 808 328 299 474

2007 8 812 852 660 772 57 458 858 382 333 444

2008 8 603 568 648 462 56 388 978 488 391 620

2009 8 749 099 626 465 54 475 993 784 444 949

2010 8 051 414 626 777 54 502 1 009 043 489 622

2011 7 402 299 700 000 55 000 1 019 755 489 482

2012 7 393 325 564 000 51 000 1 095 264 525 730

table 4.3: Sector 2 ippU: Clinker fraction for the period 2000 – 2012.

Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Country specific clinker fraction

0.76 0.75 0.73 0.73 0.73 0.72 0.73 0.71 0.7 0.71 0.69 0.69 0.69



and hydrated lime, respectively (ipCC 2006 Guidelines).


Data on glass production was obtained from the glass production industry.


For the calculation of GHG emissions from glass production, the emission factor (0.2 tonnes Co2 per tonne glass) was sourced from the 2006 ipCC Guidelines based on a typical raw material mixture, according to national glass production statistics. a typical soda-lime batch might consist of sand (56.2 weight percent), feldspar (5.3%), dolomite (9.8%), limestone (8.6%) and soda ash (20.0%). Based on this composition, one metric tonne of raw materials yields approximately 0.84 tonnes of glass, losing about 16.7% of its weight as volatiles, in this case virtually entirely Co2 (ipCC, 2006).

4.2.5 Uncertainty and time-series consistency


according to the 2006 ipCC Guidelines, if a 95% clinker fraction in portland cement is assumed then the uncertainty is in the range of 2 to 7%. Since a tier 2 method with plant–level information was used there is no need to account for imports and exports, therefore a low uncertainty (1%) was assumed.


the only available data for lime production was sourced from the SaMi report; therefore there was no comparison of data across different plants. according to the ipCC 2006 Guidelines, the uncertainty of the activity data for a tier 1 emission estimation methodology is within the range of 4 to 8%.


the only available data for glass production was sourced from the glass production industry; therefore there was no comparison of data across different plants. the uncertainty associated with use of the tier 1 emission factorandculletratio issignificantlyhighat+/-60%(ipCC, 2006, vol 3).

4.2.6 Source specific QA/QC and verification

For cement production, facility-level activity data submitted by facilities was compared with data published by the cement association as well as data reported in the SaMi reports. a comparison was made between facility-level data and the SaMi data and it was evident that there are discrepancies between the two sources of data. to give a few examples; for 2000, SaMi reported the total cement production as 3 742 126 tonnes and industry reported the production for the same year as 6 436 640 tonnes; for 2009 SaMi reported production as 5 027 293 tonnes and the industry reported production as 8 749 099 tonnes. these differences lead to increased uncertainty and the reasons for the discrepancies need to be further investigated before the next inventory. Corrections were made in facility-level data to ensure that emissions were categorised according to ipCC categorization. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvement.

4.2.7 Source-specific recalculations

No recalculations were performed for cement production, lime production or glass production.


4.2.8 Source-specificplannedimprovements and recommendations


an improvement would be the collection of activity data from all cement production plants in South africa. the activity data must include the Cao content of the clinker and the fraction of this Cao from carbonate. according to the 2006 ipCC Guidelines, it is good practice to separate Cao from non-carbonate sources (e.g. slag and flyash)andCaOcontentoftheclinkerwhencalculatingemissions. it is evident that there are discrepancies between the cement production data from industry and the cement production data published by the DMr, as a recommendation, the DMr should work with the cement production industry to ensure accuracy and consistency between the two data sources.


it is recommended that activity data be collected from all lime production plants in South africa. another improvement would be the development of country-specificemissionfactors.


Determining country-specif ic emission factors is recommended for the improvement of emission estimates from this category. one of the largest sources of uncertainty in the emissions estimate (tier 1 and tier 2) for glass production is the cullet ratio. the amount of recycled glass used can vary across facilities in a country and in the same facility over time. the cullet ratio might be a good candidate for more in-depth investigation.

4.3 Chemical Industry [2b]

this category estimates GHG emissions from the production of both organic and inorganic chemicals in South africa. the chemical industry in South africa is mainlydevelopedthroughthegasificationofcoalbecausethecountryhasnosignificantoilreserves.GHGemissionsfrom the following chemical production processes were reported: ammonia production, nitric acid production, carbide production, titanium dioxide production and carbon black. the chemical industry in South africa contributes approximately 5% to the GDp and 23% of its manufacturing sales. the chemical products in South africa can be divided into four categories: base chemicals, intermediate chemicals, chemical end-products, and speciality end-products. Chemical products include ammonia, waxes, solvents, plastics, paints, explosives and fertilizers.


4.3.1.1 Ammonia Production [2B1]

ammonia production is the most important nitrogenous material produced and is a major industrial chemical. according to the 2006 ipCC Guidelines (p.3.11), ammonia gas can be used directly as a fertilizer, in heat treating, paper pulping, nitric acid and nitrates manufacture, nitric acid ester and nitro compound manufacture, in explosives of various types and as a refrigerant.

4.3.1.2 Nitric Acid Production [2B2]

Nitric acid is a raw material used mainly in the production of nitrogenous-based fertilizer. according to the 2006 ipCC Guidelines (p.3.19), during the production of nitric acid, nitrous oxide is generated as an unintended by-product of high-temperature catalytic oxidation of ammonia.



4.3.1.3 Carbide Production [2B5]

Carbide production can result in GHG emissions such as Co2 and CH4. according to the 2006 ipCC Guidelines (p.3.39), calcium carbide is manufactured by heating calcium carbonate (limestone) and subsequently reducing Cao with carbon (e.g. petroleum coke).

4.3.1.4 Titanium Dioxide Production [2B6]

titanium dioxide (tio2) is a white pigment used mainly in paint manufacture, paper, plastics, rubber, ceramics, fabrics, floor coverings, printing ink, among others. according 2006 ipCC Guidelines (p. 3.47), there are three processes in titanium dioxide production that result in GHG emissions, namely, a) titanium slag production in electric furnaces; b) synthetic rutile production using the Becher process and c) rutile tio2 production through the chloride route.

4.3.1.5 Carbon Black [2B8F]

Carbon black is produced from petroleum-based or coal-based feed stocks using the furnace black process (ipCC, 2006). primary fossil fuels in carbon black production include natural gas, petroleum and coal. the use of these fossil fuels may involve the combustion of hydrocarbon content for heat rising and the production of secondary fuels (ipCC, 2006, p.3.56). GHG missions from the combustion of fuels obtained from feed stocks should be allocated to the source category in the ippU sector, however, where the fuels are not used within the source category but are transferred out of the process for combustion elsewhere, these emissions should be reported in the appropriate energy sector source category (ipCC, 2006, p. 3.56). Commonly, the largest percentage of carbon black is used in the tyre and rubber industry, and the rest is used as pigment in applications such as ink and carbon dry-cell batteries.

4.3.2 Trends in emissions

Thechemicalindustriessubsectorcontainsconfidentialinformation, so, following the ipCC Guidelines for reporting confidential information, no disaggregated source-category level emission data are reported; only the emissions at the sector scale are discussed. emission estimates are, however, based on bottom-up activity data and methodologies.

the chemical industries contributed emissions totalling 26 034 Gg Co2eq over the period 2000 to 2012. emissions fromthiscategoryfluctuatedconsiderablyoverthe12-year period, with a maximum of 2 889 Gg Co2eq in 2005 and a minimum of 1 011 Gg Co2eq in 2010 (Figure 4.4: Sector 2 ippU: trend and emission levels in the chemical industries between 2000 and 2012).

overall there was a 50.8% decline in GHG emissions from this category over the 12-year period, largely due to N2o emission reductions in nitric acid production.



GHG emission estimates from ammonia production were obtained through the tier 3 approach; the GHG emissions from this category were calculated based on actual process balance analysis. the emission factors will not be provided, for the reason that there is only one company that produces ammonia and therefore the total consumptionisconfidential.

4.3.3.2 Nitric Acid Production [2b2]

a tier 3 approach was used for the calculation of GHG emissions from nitric acid production, using production data and relevant emission factors. the GHG emissions in this category were calculated based on actual process balance analysis.



emission estimates for carbide production were obtained by using the tier 1 approach. Default ipCC 2006 emission factors were used. the GHG emissions from carbide production were estimated from activity data on petroleum coke consumption, which is in line with the 2006 ipCC Guidelines.


the tier 1 approach was used for calculating GHG emissions from titanium dioxide production, using 2006 ipCC default emission factors.


tier 1 was the main approach used in calculating GHG

emissions from carbon black production, using production data and relevant emission factors. ipCC 2006 default emission factors were used in all GHG emission estimations.

4.3.4 Data sources


activity data from ammonia production were not provided, but rather the total emission estimates were obtained from the ammonia production plants. the emissions were calculated based on actual process balance analysis. the total GHG emissions for 2000 were not provided for; therefore it was assumed that the GHG emissions for 2001 were similar to the GHG emissions for 2000.

Figure 4.4: Sector 2 ippU: trend and emission levels in the chemical industries between 2000 and 2012.




the amount of nitric acid emissions released was sourced from all nitric acid production plants. the GHG emissions were calculated based on continuous monitoring (tier 3 approach). Sasol emissions were also included.


the carbide production values were sourced from the carbide production plants and emissions were calculated based on actual process balance analysis.


For the calculation of GHG emissions from carbide production, the ipCC 2006 Co2 emission factor (1.090 tonnes Co2 per tonne carbide production) was used.


the titanium dioxide emissions data were sourced from the titanium dioxide production plants. emission estimates were based on a mass-balance approach.


Carbon black activity data was sourced directly from industry.


For the calculation of GHG emissions from carbon black production, the ipCC 2006 default Co2 and CH4 emission factors were used (p. 3.80). it was assumed that carbon black is produced through the furnace black process.



according to the 2006 ipCC Guidelines (p. 3.16), the plant-level activity data required for the tier 3 approach arethetotalfuelrequirementclassifiedbyfueltype;CO2 recovered for downstream use or other applications; and ammonia production. it is recommended that uncertainty estimates are obtained at the plant level, which should be lower than the uncertainty values associated with the ipCC default emission factors.


according to the 2006 ipCC Guidelines (p. 3.24) the plant-level activity data required for the tier 3 approach include production data disaggregated by technology and abatement system type. according the 2006 ipCC Guidelines (p. 3.24), default emission factors have very high uncertainties for two reasons: a) N2o may be generated in the gauze reactor section of nitric acid production as an unintended reaction by-product; and b) the exhaust gas may or may not be treated for Nox control and the Nox abatement system may or may not reduce the N2o concentration of the treated gas. the uncertainty measures of default emission factors are +/- 2%. the ipCC guidelines suggest that where uncertainty values are not available from other sources, as is the case for this inventory, this default value of ±2 percent should be applied to the activity data (ipCC 2006, vol 3, Chpt. 3, pg. 3.25). For emission factors the default uncertainty range between 10% and 40% for a tier 2 approach (ipCC 2006, vol 3, Chpt. 3, and pg. 3.23, table 3.3). Since a tier 3 approach was applied in this inventory the lower uncertainty value of 10% was assumed.

4.3.5.3 Carbide Production [2b5]

the total GHG emissions were sourced from the specificcarbideproductionplantsthereforetherewas


no comparison of data across different plants. the default emission factors are generally uncertain because industrial-scale carbide production processes differ from the stoichiometry of theoretical chemical reactions (ipCC, 2006, p. 3.45). according to the ipCC 2006 Guidelines (p. 3.45), the uncertainty of the activity data that accompanies the method used here is approximately 10%.


the total GHG emissions were sourced from the specifictitaniumdioxideproductionplantstherefore,no comparison of data across different plants was made. according to the ipCC 2006 Guidelines (p. 3.50), the uncertainty of the activity data that accompanies the method used here is approximately 5%.


the activity data was sourced from disaggregated national totals; therefore QC measures were not applied. according to the ipCC 2006 Guidelines, the uncertainty of the activity data that accompanies the method used here is in the range of -15% to +15% for Co2 emission factors and between -85% to +85% for CH4 emission factors.

4.3.6 Source-specific QA/QC and verification

4.3.6.1 Chemical Industry [2B]

No source-specific Qa/QC was performed for the chemical industry. However, activity data and material balancedatawasverifiedwithindustrysectors.


No recalculations were performed.

4.3.8 Source-specific planned improvements and recommendations

For ammonia, nitric acid, carbide and titanium production itisrecommendedthatthecountry-specificemissionfactors which were applied by the industry be made transparent.Thiswouldallowefficientqualityassuranceand control of the emission factors used. there is high uncertainty on many of the default emission factors, so havingcountry-specificemissionfactorswouldreducethisuncertainty.Developmentofcountry-specificemissionfactors and/or moving to a material approach will improve emission estimates for the carbon black source category.

4.4 Metal industry [2C]

this subcategory relates to emissions resulting from the production of metals. processes covered for this inventory report include the production of iron and steel, ferroalloys, aluminium, lead, and zinc. estimates were made for emissions of Co2 from the manufacture of all the metals, CH4 from ferroalloy production, andperfluorocarbons(CF4 and C2F6) from aluminium production.


4.4.1.1 Iron and Steel Production [2C1]

iron and steel production results in the emission of Co2, CH4 and N2o. according to the 2006 ipCC Guidelines (p. 4.9), the iron and steel industry broadly consists of primary facilities that produce both iron and steel; secondary steel-making facilities; iron production facilities; and offsite production of metallurgical coke. according to the World Steel association (2010), South africa is the 21st-largest crude steel producer in the world. the range of primary steelproductsandsemi-finishedproductsmanufacturedin South africa includes: billets; blooms; slabs; forgings;



light-, medium- and heavy sections and bars; reinforcing bar; railway track material; wire rod; seamless tubes; plates; hot- and cold-rolled coils and sheets; electrolytic galvanised coils and sheets; tinplate; and pre-painted coils and sheets. the range of primary stainless steel products andsemi-finishedproductsmanufacturedinSouthAfricainclude slabs, plates, and hot- and cold-rolled coils and sheets.

4.4.1.2 Ferroalloy Production [2C2]

Ferroalloy refers to concentrated alloys of iron and one or more metals such as silicon, manganese, chromium, molybdenum, vanadium and tungsten. Ferroalloy plants manufacture concentrated compounds that are delivered to steel production plants to be incorporated in alloy steels. Ferroalloy production involves a metallurgical reduction processthatresultsinsignificantcarbondioxideemissions(ipCC, 2006, p. 4.32). South africa is the world’s largest producer of chromium and vanadium ores, and the leading supplier of their alloys (SaMi, 2013). South africa is also the largest producer of iron and manganese ores, and an important supplier of ferromanganese, ferrosilicon and silicon metal (SaMi, 2013).

4.4.1.3 Aluminium Production [2C3]

according to the 2006 ipCC Guidelines, aluminium production is realised via the Hall-Heroult electrolytic process. in this process, electrolytic reduction cells differ intheformandconfigurationofthecarbonanodeandalumina feed system.

Themostsignificantprocessemissionsare(IPCC,2006,p. 4.43):

• Carbon dioxide (Co2) emissions from the consumption of carbon anodes in the reaction to convert aluminium oxide to aluminium metal;

• Perfluorocarbon(PFC)emissionsofCF4andC2F6 during anode effects. also emitted are smaller amounts of process emissions, Co, So2, and

NMvoCs. SF6 is not emitted during the electrolytic process and is only rarely used in the aluminium manufacturing process, where small quantities are emittedwhenfluxingspecializedhigh-magnesiumaluminium alloys.

4.4.1.4 Lead Production [2C5]

according to the 2006 ipCC Guidelines, there are two primary processes for the production of lead bullion from lead concentrates:

• Sintering/smelting, which consists of sequential sintering and smelting steps and constitutes approximately 7% of the primary production; and

• Direct smelting, which eliminates the sintering step and constitutes 22% of primary lead production.

4.4.1.5 Zinc Production [2C6]

according to the 2006 ipCC Guidelines, there are three primary processes for the production of zinc:

• electro-thermic distillation; this is a metallurgical process that combines roasted concentrate and secondary zinc products into sinter that is combusted to remove zinc, halides, cadmium and other impurities. the reduction results in the release of non-energy Co2 emissions.

• the pyrometallurgical process: this involves the utilization of an imperial Smelting Furnace, which allows for the simultaneous treatment of zinc and zinc concentrates. the process results in the simultaneous production of lead and zinc and the release of non-energy Co2 emissions.

• the electrolytic: this is a hydrometallurgical technique, during which zinc sulphide is calcinated, resulting in the production of zinc oxide. the process does not result in non-energy Co2 emissions.



emissions from the metal industry totalled 367 785 Gg Co2eq between 2000 and 2012. the major contributor over this period was iron and steel production (55.4 %), followed by ferroalloy production (36.0 %) and aluminium production (8.0 %) (Figure 4.5). in 2000 almost half (47.1%) of the total GHG emissions (in Gg Co2eq) from aluminium production were pFC emissions. this rose to 65.0% in 2011 and 2012 due to the closure of the Soderberg and Side-Worked pre-Bake processes in 2009. the aluminium plants released large amounts of C2F4 and CF4 during 2011and2012duetoinefficientoperations(switchingon and off at short notice) as they were used to control the electricity grid.

total emissions in the metal industry increased between

2000 and 2006, with a sharp decline (14.2 %) between 2006 and 2009 (Figure 4.6). this decrease was evident in the three major contributing industries, with emissions from iron and steel production showing a decline of 25.7%, aluminium production emissions declined by 39.6% and zinc production emissions declined by 4.4% over this period. Ferroalloy industry emissions increased in most years, except in 2008 and 2012, when they declined by 0.6% and 5.0% respectively. these declines could be attributed to reduced production caused by electricity supply challenges and decreased demand following the economic crisis that occurred during 2008/2009. Between 2009 and 2012 total GHG emissions from the metal industries increased by 16.2%, from 27 328 Gg Co2eq in 2009 to 29 721 Gg Co2eq in 2012. although emissions from zinc production were relatively small, emissions decreased by 100% between 2009 and 2012 due to the closure of zinc production facilities in South africa.

Figure 4.5: Sector 2 ippU: trend and emission levels in the chemical industries between 2000 and 2012.



During 2003/04 South africa’s lead mine production declined by 6.2%, as did the emissions, due mainly to the depletion of a part of the Broken Hill ore body at Black Mountain mine, which contained a higher-grade ore (DMr, 2005). During 2004/05 zinc production decreased by 6.3% due to the closure of Metorex’s Maranda operation in July 2004 (DMr, 2004) and emissions declined by 1.0% over this period. in 2009, GHG emissions from zinc production increased by 4.9%, and this was attributed to new mine developments, such as the pering Mine and the anglo american Black Mountain mine and Gamsberg project (DMr, 2009).

over the 12-year period emissions from iron and steel, lead and zinc production declined (8.5%, 30.3% and 100%, respectively), while emissions increased by 45.9% and 43.9% from the aluminium and ferroalloy industries, respectively.

4.4.3 Methodology

4.4.3.1 Iron and Steel Production [2C1]

a combination of the tier 1 and tier 2 approaches (country-specif ic emission factors) was applied to calculate the GHG emissions from iron and steel for the different process types. Default ipCC emission factors were used for the calculation of GHG emissions from basic oxygen furnace, electric arc furnace and pig iron production,andcountry-specificemissionfactorswereused for the estimation of GHG emissions from direct reduced iron production. the separation of energy and process emissions emanating from the use of coke was not done due to a lack of disaggregated information on coke consumption. Hence, energy-related emissions from iron and steel production have been accounted for through the application of default ipCC emission factors.

Figure 4.6: Sector 2 ippU: trend and emission levels in the metal industry, 2000 – 2012.


tabl

e 4.

4:

Sect

or 2

ippU

: act

ivit

y da

ta fo

r th

e va

riou

s m

etal

indu

stri

es, 2

000

– 20

12.

Con

sum

ptio

n (t

ons)

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

Iron

and

st

eel

prod

ucti

on

Basi

c ox

ygen

fu

rnac

e pr

oduc

tion

4 67

4 51

14

849

655

5 05

1 93

65

083

168

4 94

9 69

35

255

831

5 17

3 67

64

521

461

4 50

4 27

53

953

709

4 36

6 72

73

991

686

3 90

4 27

6

elec

tric

arc

furn

ace

4 54

9 82

84

716

954

4 88

8 87

05

353

456

5 50

8 48

85

089

818

5 41

3 20

45

473

908

4 58

1 52

34

359

556

4 23

5 99

33

554

803

3 90

4 27

6

pig

iron

prod

uctio

n (n

ot c

onve

rted

into

st

eel)

4 67

4 51

14

849

655

5 05

1 93

64

474

699

4 22

4 48

74

441

904

4 43

5 55

13

642

520

3 74

6 78

63

184

566

3 69

5 32

74

603

558

4 59

9 01

5

Dire

ct r

educ

ed

iron

(Dr

i) pr

oduc

tion

1 55

2 55

31

220

890

1 34

0 97

61

542

008

1 63

2 76

71

781

108

1 75

3 58

51

735

914

1 17

7 92

51

339

720

1 12

0 45

21

414

164

1 49

3 42

0

oth

er (

Cor

ex e

tc.)

705

872

706

225

706

578

706

931

733

761

735

378

739

818

705

428

460

746

429

916

584

452

570

129

677

891

Ferr

oallo

y pr

oduc

tion

Chr

omiu

m a

lloys

2 57

4 00

02

141

000

2 35

1 00

02

813

000

3 03

2 00

02

802

000

3 03

0 00

03

561

000

3 26

9 00

02

346

000

3 60

7 00

03

422

000

3 06

3 00

0

Man

gane

se a

lloys

(7

% C

)59

6 87

352

3 84

461

8 95

460

7 36

261

1 91

457

0 57

465

6 23

569

8 65

450

2 63

127

4 92

347

3 00

071

4 00

070

6 00

0

Man

gane

se a

lloys

(1

% C

)31

0 40

025

9 17

631

5 80

231

3 15

237

3 92

827

5 32

427

7 70

332

7 79

425

9 01

411

7 68

331

7 00

035

0 00

017

7 00

0

Silic

on a

lloys

(a

ssum

e 65

% S

i)10

8 50

010

7 60

014

1 70

013

5 30

014

0 60

012

7 00

014

8 90

013

9 60

013

4 50

011

0 40

012

8 76

012

6 00

0 8

3 00

0

Silic

on m

etal

40 6

0039

400

42 5

0048

500

50 5

0053

500

53 3

0050

300

51 8

0038

600

46 4

0058

800

53 0

00

Alu

min

ium

pr

oduc

tion

preb

ake

586

868

573

285

623

778

629

668

778

067

784

638

800

668

808

630

788

859

811

324

808

795

812

209

666

790

Sode

rber

g89

572

85 9

7382

749

89 0

3788

644

86 5

2990

082

91 3

3422

722

00

00

CW

pB58

6 86

857

3 28

562

3 77

862

9 66

877

8 06

778

4 63

880

0 66

880

8 63

078

8 85

981

1 32

480

8 79

581

2 20

966

6 79

0

SWpB

9 98

09

763

9 19

210

084

9 78

410

002

9 97

49

925

804

00

00

vSS

79 5

9276

210

73 5

5778

952

78 8

6076

527

80 1

0881

409

21 9

170

00

0

Lead

pr

oduc

tion

lead

75 3

0051

800

49 4

0039

900

37 5

0042

200

48 3

0041

900

46 4

0049

100

50 6

0054

460

52 4

89

Zin

c pr

oduc

tion

Zin

c11

3 00

011

0 00

011

3 00

011

3 00

010

5 00

010

4 00

090

000

101

000

82 0

0086

000

85 1

0072

000

0



4.4.3.2 Ferroalloy Production [2C2]

Ferrochromium production emissions are based on plant-level data (tier 3 method), while the rest of the Ferroalloys are based on t1 approach.

4.4.3.3 Aluminium Production [2C3]

a tier 3 methodology where the amount of CF4 and C2F6 produced are tracked were used to determine emissions in this category.

4.4.3.4 Lead Production [2C5]

emissions from lead production were estimated using a tier 1 approach. it was assumed that that lead production was 80% imperial Smelting Furnace and 20% direct smelting.

4.4.3.5 Zinc Production [2C6]

emissions from zinc production were calculated with the tier 1 approach.

4.4.4 Data sources

Metal industry emission estimates were based on data from two main sources: the SaiSi (2008) provided data for iron and steel production, while production data for all the other metals were obtained from the SaMi report (Department of Minerals and energy, 2013).

4.4.4.1 Iron and Steel Production[2C1]

the SaiSi provided data for iron and steel production (Table 4.4).


a combination of country-specific emission factors and ipCC default emission factors were applied for

the calculation of GHG emissions from iron and steel. Country-specificemissionfactorsweresourcedfromoneof the iron and steel companies in South africa (Table 4.5) and these were based on actual process analysis attherespectiveplants.Thecountry-specificemissionfactor for electric arc furnace (eaF) production is slightly higher than the ipCC default value; this emission factor was, however, not used for the estimation of GHG emissions from eaF because was based on a small sample and needs further investigation before it can be applied.Thecountry-specificemissionfactorforDirectreduced iron (Dri) production is more than twice the default factor (Table 4.5).Thiscountryspecificfactorwas used for estimating GHG emissions as it was based on a comprehensive carbon balance analysis. Differences in feedstock material and origin results in higher emission factors compared with the ipCC default emission factor values, which assume consistent feedstock conditions across countries. the Other category values were based on production by the Corex process. this process is 50% Basic oxygen Furnace and 50% electric arch Furnace, therefore, a weighted emission factor (0.77 t Co2/ t production) accounting for these two processes was applied to the Other category.

Table4.5: Sector2IPPU:Comparisonofthecountry-specificemission factors for iron and steel production and the ipCC 2006 default values (Source: iron and steel Companies; ipCC 2006 Guidelines).

Type of Technology

CO2 Emission factor

(tonne CO2 / tonne iron and steel)

Country-specific

ipCC 2006 default

Basic oxygen Furnace 1.46

electric arc Furnace 1.1 0.08

pig iron production (not converted into steel)

- 1.35

Direct reduced iron (Dri) production

1.525 0.7

Sinter production 0.34 0.2

Con

sum

ptio

n (t

ons)

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

Iron

and

st

eel

prod

ucti

on

Basi

c ox

ygen

fu

rnac

e pr

oduc

tion

4 67

4 51

14

849

655

5 05

1 93

65

083

168

4 94

9 69

35

255

831

5 17

3 67

64

521

461

4 50

4 27

53

953

709

4 36

6 72

73

991

686

3 90

4 27

6

elec

tric

arc

furn

ace

4 54

9 82

84

716

954

4 88

8 87

05

353

456

5 50

8 48

85

089

818

5 41

3 20

45

473

908

4 58

1 52

34

359

556

4 23

5 99

33

554

803

3 90

4 27

6

pig

iron

prod

uctio

n (n

ot c

onve

rted

into

st

eel)

4 67

4 51

14

849

655

5 05

1 93

64

474

699

4 22

4 48

74

441

904

4 43

5 55

13

642

520

3 74

6 78

63

184

566

3 69

5 32

74

603

558

4 59

9 01

5

Dire

ct r

educ

ed

iron

(Dr

i) pr

oduc

tion

1 55

2 55

31

220

890

1 34

0 97

61

542

008

1 63

2 76

71

781

108

1 75

3 58

51

735

914

1 17

7 92

51

339

720

1 12

0 45

21

414

164

1 49

3 42

0

oth

er (

Cor

ex e

tc.)

705

872

706

225

706

578

706

931

733

761

735

378

739

818

705

428

460

746

429

916

584

452

570

129

677

891

Ferr

oallo

y pr

oduc

tion

Chr

omiu

m a

lloys

2 57

4 00

02

141

000

2 35

1 00

02

813

000

3 03

2 00

02

802

000

3 03

0 00

03

561

000

3 26

9 00

02

346

000

3 60

7 00

03

422

000

3 06

3 00

0

Man

gane

se a

lloys

(7

% C

)59

6 87

352

3 84

461

8 95

460

7 36

261

1 91

457

0 57

465

6 23

569

8 65

450

2 63

127

4 92

347

3 00

071

4 00

070

6 00

0

Man

gane

se a

lloys

(1

% C

)31

0 40

025

9 17

631

5 80

231

3 15

237

3 92

827

5 32

427

7 70

332

7 79

425

9 01

411

7 68

331

7 00

035

0 00

017

7 00

0

Silic

on a

lloys

(a

ssum

e 65

% S

i)10

8 50

010

7 60

014

1 70

013

5 30

014

0 60

012

7 00

014

8 90

013

9 60

013

4 50

011

0 40

012

8 76

012

6 00

0 8

3 00

0

Silic

on m

etal

40 6

0039

400

42 5

0048

500

50 5

0053

500

53 3

0050

300

51 8

0038

600

46 4

0058

800

53 0

00

Alu

min

ium

pr

oduc

tion

preb

ake

586

868

573

285

623

778

629

668

778

067

784

638

800

668

808

630

788

859

811

324

808

795

812

209

666

790

Sode

rber

g89

572

85 9

7382

749

89 0

3788

644

86 5

2990

082

91 3

3422

722

00

00

CW

pB58

6 86

857

3 28

562

3 77

862

9 66

877

8 06

778

4 63

880

0 66

880

8 63

078

8 85

981

1 32

480

8 79

581

2 20

966

6 79

0

SWpB

9 98

09

763

9 19

210

084

9 78

410

002

9 97

49

925

804

00

00

vSS

79 5

9276

210

73 5

5778

952

78 8

6076

527

80 1

0881

409

21 9

170

00

0

Lead

pr

oduc

tion

lead

75 3

0051

800

49 4

0039

900

37 5

0042

200

48 3

0041

900

46 4

0049

100

50 6

0054

460

52 4

89

Zin

c pr

oduc

tion

Zin

c11

3 00

011

0 00

011

3 00

011

3 00

010

5 00

010

4 00

090

000

101

000

82 0

0086

000

85 1

0072

000

0


4.4.4.2 Ferroalloy, Aluminium, Lead and Zinc Production

the source of activity data for these categories was sourced from the SaMi report (Table 4.4). For ferromanganese production the 7% C values were taken to be the high and medium carbon ferromanganese and the 1% C values were the other manganese alloys (DMr, 2014; tables 77 and 78). ipCC default emission factors were used for ferroalloy production (Table 4.6.), except for ferrochromium production where a tier 3 approach was used. a tier 3 method was also applied for aluminium production. the emission factor of 0.52 tonnes Co2 per tonne lead and 1.72 tonnes Co2 per tonne zinc were sourced from the ipCC 2006 Guidelines.


the necessary quality-control measures were used to minimise estimation errors. the tier 1 approach for metal production emission estimates generates a number of uncertainties. For example, the ipCC 2006 Guidelines explain that applying tier 1 to default emission factors for iron and steel production may have an uncertainty

of ± 25% (ipCC 2006, vol 3, Chpt 4, pg. 4.40, table 4.9). the same range of uncertainty is associated with the tier 1 approach for ferroalloy production emission factors. For this inventory the maximum default uncertainty for t1 of 25% was assumed for the eF. there is a default 5% uncertainty on the activity data (ipCC 2006, table 4.9). even though a tier 3 approach was used for aluminium production emission no data was collected on uncertainty. the default uncertainty for CF4 and C2F6 range from -99% to +380% (ipCC 2006, vol 3, Chpt 4, page 4.54, table 4.15).Sinceacountryspecificapproachwasusedtheuncertainty is expected to be lower than the tier 1 default uncertainties, therefore 190% (half the default uncertainty) was applied for the uncertainty on these emission factors.


an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvement.


recalculation were completed for all years between 2000 and 2012 due to the following improvements and updates:

• Zinc production:

o Updated zinc production data between 2004 and 2010;

• Ferroalloys:

o Updated ferromanganese activity data for 2010;

• Steel and iron production:

o emission factor for eaF was stated to be 2.1 tonnes Co2/tonne product, but the source ofthedatacouldnotbeverifiedsotheIPCCdefault factor (0.08 tonnes Co2/tonne product) was used instead;

o Due to changes in the eaF emission factor

table 4.6: Sector 2 ippU: emission factors for ferroalloy production (Source: 2006 ipCC Guidelines).

Ferroalloy Type

Co2 CH4

(tonnes per ferroalloys tonne production)

Ferrosilicon (45%) Si 2.5 n/a

Ferrosilicon (65%) Si 3.6 1

Ferrosilicon (75%) Si 4 1

Ferrosilicon (90%) Si 4.8 1.1

Ferromanganese (7% C) 1.3 n/a

Ferromanganese (1% C) 1.5 n/a

Silicomanganese 1.4 n/a

Silicon metal 5 1.2



so the emission factor for the Corex process (listed as other technologies in the steel and iron production category) was also adjusted; and

o emission factor for pig iron reduction was corrected to the ipCC default value (1.35 tonne Co2/tonne product) as the value of 1.65 tonnes Co2/tonneproductcouldnotbeverified.


as with most other subcategories, completeness of data is urgently needed for metal production activities. other improvements would be ensuring that accurate activity dataarecollectedandcountryspecificemissionfactorsare determined. in order to reduce uncertainty in the Ferroalloy production emissions it is recommended thatsitespecificdataisurgentlyacquired.Forreducinguncertainty in the cement production emissions it is important for the DMr to work together with the cement production industry to ensure consistent reporting of data.

4.5 Non-energy use of fuels and solvent use [2D]

4.5.1 Source-category description

Non-energy use of fuels and solvents includes lubricants and paraffin wax. the use of solvents can result in evaporative emissions of various NMvoCs, which can be oxidized and released into the atmosphere. according to the 2006 ipCC Guidelines (p. 5.16), white spirit is used as an extraction solvent, cleaning solvent, degreasing solvent and as a solvent in aerosols, paints, wood preservatives, varnishes and asphalt products. lubricants are used in industrial and transport applications. lubricants are divided into two types, namely, motor and industrial oils, andgreasesthatdifferinphysicalcharacteristics.Paraffin

waxisusedinproductssuchaspetroleumjelly,paraffinwaxesandotherwaxes(saturatedhydrocarbons).Paraffinwaxes are used in applications such as candles, corrugated boxes, paper coating, board sizing, food production, wax polishes, surfactants (as used in detergents) and many others (ipCC, 2006, p.5.11).


total GHG emissions from the non-energy products from fuels and solvent usecategoryfluctuatedbetween196GgCo2eq and 246 Gg Co2eq between 2000 and 2004, and hovered around 230 Gg Co2eq between 2007 and 2010, with a peak in emissions (504 Gg Co2eq) occurring in 2006 (Figure 4.7).

emissions from lubricant use contributed between 93.5% and 99.1% of the total emissions from this category.


emissions for this category were estimated using a tier 1 approach. in line with the 2006 ipCC Guidelines (p.5.9), it was assumed that 90% of the mass of lubricants is oil and 10% is grease.

4.5.4 Data sources

the source of activity data for solvents was the energy balance tables published annually by the Doe (Department of energy (Doe), 2008. energy Balances Statistics. DepartmentofEnergy,http://www.energy.gov.za/files/energyStats_frame.html. [accessed November 2012]).

4.5.4.1 Emission Factors

the ipCC 2006 default emission factor for lubricating oils, grease and lubricants (0.2 tonnes Co2 per tJ product) was used in the calculation of emissions from lubricant andparaffinwaxuse.



the default oxidised during use (oDU) factors available in the ipCC guidelines are very uncertain, as they are based on limited knowledge of typical lubricant oxidation rates. expert judgment suggests using a default uncertainty of50%.Thecarboncontentcoefficientsarebasedontwo studies of the carbon content and heating value of lubricants, from which an uncertainty range of about ±3 % was estimated (ipCC, 2006). according to the ipCC guidelines much of the uncertainty in emission estimates isrelatedtothedifficultyindeterminingthequantityof non-energy products used in individual countries. For this a default of 5% may be used in countries with well-developed energy statistics and 10 to 20 % in other countries, based on expert judgement of the accuracy of energy statistics.

table 4.7: Sector 2 ippU: total fuel consumption in the non-energy use of fuels and solvent use category, 2000 – 2010.

PeriodFuel consumption

lubricants (tJ) ParaffinWax(TJ)

2000 12 851 507

2001 15 092 314

2002 16 561 506

2003 16 430 521

2004 16 295 490

2005 31 549 350

2006 34 391 324

2007 15 819 141

2008 14 891 182

2009 15 707 231

2010 15 715 231

2011 13 130 260

2012 17 085 1 188

Figure 4.7: Sector 2 ippU: trend and emission levels from the non-energy products from fuels and solvent use category, 2000 – 2012.




an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and to identify areas of improvement.


Nosource-specificrecalculationswereperformed.


energy balances remain the source of activity data for thissourcecategoryand,therefore,nosource-specificimprovements are planned in the future. However, improvements in data collection for energy balances would reduce uncertainty in fuel use data.

4.6 Production uses as substitutes for ozone-depleting substances [2F]

the Montreal protocol on Substances that Deplete the ozone layer (a protocol to the vienna Convention for the protection of the ozone layer) is an international treaty designed to protect the ozone layer by phasing out the production of numerous substances believed to beresponsibleforozonedepletion.Hydrofluorocarbons(HFCs)and,toalimitedextent,perfluorocarbons(PFCs)are serving as alternatives to ozone-depleting substances (oDS) being phased out under this protocol. according to the 2006 ipCC Guidelines, current application areas of HFCs and pFCs include refrigeration and air conditioning; fire suppression and explosion protection; aerosols; solvent cleaning; foam blowing; and other applications (equipment sterilisation, tobacco expansion applications,

and as solvents in the manufacture of adhesives, coatings and inks).


total Co2eq emissions from HFC’s increased from 842 Gg Co2eq in 2005 to 2 096 Gg Co2eq in 2010, then declined again to 1 396 in 2012 (Figure 4.8). there was no available data for the years prior to 2005. HFC-134a is the biggest contributor (average contribution of 80.8%) and HFC-143a contributing an average of 8.9%.


the tier 1 approach was used to estimate emissions from substitutes for ozone-depleting substances. the calculation of GHG emissions was done through a calculator provided by the ipCC. it was assumed that the equipment lifespan was 15 years and the emission factor from the installed base was 15%. these assumptions were based on the defaults from the 2006 ipCC Guidelines.

4.6.3 Data sources

activity data for oDS substitutes were sourced from the oDS database managed by the Dea. the oDS database registers imports and exports of oDS substances and their replacements. the activity data has been updated and improvedsincethelastinventory.Thisdataissufficientto follow the tier 1 methodology which requires national statisticsoninflowsandoutflowsofODSreplacementsubstances.


there may be a wide range of other applications, therefore, it is not possible to give default uncertainties for these sources. However, procedures should be put in


place to assess levels of uncertainty in accordance with the practices outlined in volume 1 Chapter 3 of the ipCC Guidelines.


Source-specific quality control is performed by the oDS database manager. the data submitted by oDS distributorswasconfirmedwiththeInternationalTradeadministration Commission (itaC), which is responsible for issuing import/export permits. in addition, the yearly import/exportODSfigureswerecomparedwithimport/export data obtained from the South african revenue Service (SarS).


Due to updated activity data recalculations were completed for this category for all years between 2000 and 2012.


according to the ipCC Guidelines, it is good practice to estimate emissions from oDS replacement use by end-use sector (e.g., foam blowing, refrigeration etc.). Future improvements would involve the collection of data at the sector level such that default emission factors can be applied at sectoral level.

Figure 4.8: Sector 2 ippU: trends and emission levels of HFC’s, 2005 – 2012.



5.1 Overview of the sector

this section includes GHG emissions and removals from agriculture as well as land use and forestry. Based on the ipCC 2006 Guidelines, the following categories are included in the emission estimates:

• livestock

o enteric fermentation (ipCC Section 3a1)

o Manure management (ipCC Section 3a2)

• land

o Forest land (ipCC Section 3B1)

o Cropland (ipCC Section 3B2)

o Grassland (ipCC Section 3B3)

o Wetlands (ipCC Section 3B4)

o Settlements (ipCC Section 3B5)

o other land (ipCC Section 3B6)

• aggregate sources and non-Co2 emissions on land

o Biomass burning (ipCC Section 3C1)

o liming (ipCC Section 3C2)

o Urea application (ipCC Section 3C3)

o Direct N2o emission from managed soils (ipCC Section 3C4)

o indirect N2o emission from managed soils (ipCC Section 3C5)

o indirect N2o emission from manure management (ipCC Section 3C6)

• other

o Harvested wood products (ipCC Section 3D1)

emissions from fuel combustion in this sector are not included here as these fall under the agriculture/forestry/fisheries subsector (see Section 3.3.6.1.3) in the energy sector. Categories not included in this report are rice cultivation (3C7), and other (3C8, 3D2), as they are not applicable to South africa. the land use component includes land remaining in the same land use as well as land converted to another land use. this section includes a tier 1 approach to the mineral soil carbon pool, while organic soils are not reported on as the area of organic soilsinSouthAfricawasestimatedtobeinsignificant.

emissions from ruminants in privately owned game parks has been included as these are suggested to be managed lands as the game are fed. Game in national parks are not included as they are considered unmanaged.

Manuremanagementincludesallemissionsfromconfined,managed animal waste systems. Methane emissions from livestockmanureproducedinthefieldduringgrazingare included under manure management (3a2); however, the N2o emissions from this source are included under category 3C4 direct N2O emissions from managed soils. this is in accordance with ipCC 2006 Guidelines. Methane emissions from managed soils are regarded as non-anthropogenic and are, according to the guidelines, not included.

losses of Co2 emissions from biomass burning of forest land are included under lossesduetodisturbance in the forest land section (3B1) and not in the biomass burning (3C1) section. Section 3C1 deals with non-Co2 emissions from biomass burning in all land use types.

5. AGRICULTURE, FORESTRY AND OTHER LAND USE


5.2 GHG emissions from the AFOLU sector


the aFolU sector was a source of Co2 (Table 5.1). Thesourcefluctuatedoverthe12yearperiod,butoverallthere appeared to be an increase in the sink. total GHG emissions from livestock declined by 5.6%, from 31 162 Gg Co2eq in 2000 to 29 413 Gg Co2eq in 2012. the decline was attributed mainly to the decreasing cattle, sheep and goat populations. the land component is estimated to be a sink, varying between 4 651 Gg Co2eq and 20 570 Gg Co2eq (Table 5.1). the variation was caused mainly by changes in carbon stocks in forest lands due tovariationincarbonlosses(fireandfuelwoodlosses).

emissions from aggregated and non-CO2 emission sources varied by a maximum of 9.1% over the 12 year period. the fluctuationsinthiscategoryaredrivenmainlybychangesin biomass burning, liming and urea application. HWp estimates indicate that this subsector is a small sink of Co2, varying between a minimum of 312 Gg Co2eq (2000) and a maximum of 1 187 Gg Co2eq (2004). in 2011 this subsector was a small source of Co2. as with the land and aggregated and non-Co2 emission sources, the sink does not show any trend.

the recalculated 2010 estimate in this inventory was 49.6% higher than the previous inventory estimate of 25 713 Gg Co2eq (Department of environmental affairs, 2014). the change was attributed mainly to the use of updated land cover change maps and corrected HWp estimates. the HWp estimates were reduced by an average of 94% compared with the 2010 inventory due

table 5.1: Sector 3 - aFolU: trends in emissions and removals (Gg Co2eq) from the aFolU sector, 2000 – 2012.

3 - Agriculture, Forestry, and

Other Land Use

3.A - Livestock

3.b - Land3.C - Aggregate sources and non-CO2 emissions

sources on land

3.D - Other (HWp)

2000 45 860 31 162 -8 521 23 527 -312

2001 45 813 31 003 -8 205 23 683 -675

2002 43 971 30 771 -10 597 24 607 -817

2003 42 342 30 194 -11 777 22 951 -927

2004 44 066 30 039 -7 701 22 980 -1 185

2005 44 047 29 998 -8 561 22 832 -197

2006 41 952 29 956 -10 285 23 186 -882

2007 39 758 29 325 -11 582 22 600 -581

2008 47 225 29 969 -5 259 23 327 -781

2009 39 011 29 412 -13 019 22 738 -98

2010 38 456 30 245 -14 870 23 578 -490

2011 38 376 30 349 -15 779 23 738 82

2012 31 128 29 413 -20 346 22 530 -512

5. agriculture, Forestry and other land Use


to corrections and improved methodology.

in all years CH4 emissions contributed the most to the total aFolU emissions (Figure 5.1: Sector 3 aFolU: percentage contribution of the various GHG to the total aFolU inventory, 2000 – 2012.), but this contribution declined from 46.32% in 2000 to 38.5% in 2012. Enteric fermentation contributed an average of 92.4% of the CH4. the contribution from N2Ofluctuatedannuallybutwasinthe range of 32.5% to 38.9% with an average of 68.8% of this coming from direct N2O emissions from managed soils.

5.2.2 Key sources

a level and trend key category analysis including the aFolU sector was carried out on the 2012 data (Appendix b). Both of these assessments showed that the main key categories in this sector are land converted to forest land (3.B.1.b), enteric fermentation (from cattle (3.a.1.a)), direct N2O from managed soils – urine and dung deposits in PRP (3.C.4), land converted to grasslands (3.B.2.b) and land converted to croplands (3.B.2.b) (Table 1.10, Table 1.12). it should be noted that the key category assessment is incomplete due to the incomplete reporting of all sources.

Figure 5.1: Sector 3 aFolU: percentage contribution of the various GHG to the total aFolU inventory, 2000 – 2012.


5.2.3 Recalculations

in the livestock subsector, minor corrections were made to the beef and dairy cattle population numbers, as well as the poultry population numbers. also emissions from game on privately owned game parks were included using activity data and emission factors supplied by Du toit et al. (2013d). Due to these updated activity data and emission factors the annual livestock emissions since 2000 were recalculated.

in the land category there were several updates:

• New land-use change maps were incorporated (for 1990 and 2013/2014);

• Biomass data was updated;

• DoM was incorporated for forest lands;

• estimates from other lands were included;

• Changes in soil carbon methodology to include soil type; and

• Corrected and updated HWp methodology.

these recalculations were completed for all yeas since 2000 and all updates and improvements are discussed in the sections below.

5.3 Livestock [3A]


the GHG emissions from livestock produced a total of 29 413 Gg Co2eq in 2012. this was a 5.6% decline from the 2000 emissions (31 162 Gg Co2eq) (Figure 5.2: Sector 3 aFolU: trends in GHG emissions from the livestock category, 2000 – 2012.).

the total accumulated emissions since 2000 stand at 332

075 Gg Co2eq. emissions declined annually between 2000 and 2007, and increased by 2.2% in 2008. there was also an increase of 3.0% between 2009 and 2011. enteric fermentation accounted for an average of 93.3% of the GHG emissions from livestock, while the rest was from manure management. emissions from manure management showed an increase in emissions of 18.2% between 2000 (1 872 Gg Co2eq) and 2012 (2 214 Gg Co2eq). this increase was mainly due to the increase in the poultry population with all of the manure from this livestock category being managed. if poultry manure is excluded then there is a slight decline in manure emissions from all other livestock, with little annual variation.

5.3.2 Overview of trends in activity data

the total annual livestock population increased by 29.8% between 2000 and 2012, mainly due to growth in the poultry industry. if poultry numbers are excluded, then there was a general decline (of 6.1%) in the remaining livestock population over the 12-year period (Figure 5.3: Sector 3 aFolU - livestock: trends in livestock population numbers, 2000 – 2012.).

Thecommercialbeefcattlepopulationfluctuatedslightlyaround an average of 7 million (with a low of 6.85 million in 2002, and a high of 7.34 million in 2001), while the communalbeefpopulationfluctuatedaround5.4million(± 0.27 million). the dairy cattle population declined between 2000 and 2004 (25.6%), increased between 2004 and 2010 (31.4%), and declined by 7.5% between 2010 and 2012. Sheep and goat populations declined by 9.2% and 13.9%, respectively, and swine by 4.1% between 2000 and 2012. the annual average poultry population increased by 48.4%, from 100 million in 2000 to 149 million in 2012 (Figure 5.3: Sector 3 aFolU - livestock: trends in livestock population numbers, 2000 – 2012.).

Generally, the commercial population numbers were higher than the communal numbers, except for goats, where the communal goat population was approximately twice that of the commercial population. the game



Figure 5.3: Sector 3 aFolU - livestock: trends in livestock population numbers, 2000 – 2012.

Figure 5.2: Sector 3 aFolU: trends in GHG emissions from the livestock category, 2000 – 2012.


population in privately owned game parks was estimated at 2 285 000 (Du toit et al., 2013d). this was kept constant each year due to a lack of data.

5.3.3 Enteric fermentation [3A1]

5.3.3.1 Source-category Description

Methane is the main greenhouse gas produced from agricultural livestock production. CH4 from enteric fermentation is produced in herbivores as a by-product of the digestive process, and is released mainly through eructation and normal respiration, and a small quantity asflatus(Bullet al., 2005; Chhabra et al., 2009; Baggot et al., 2006; ipCC, 1997; ipCC, 2006). the amount of CH4 produced and emitted by an individual animal depends primarily on the animal’s digestive system and the amount and type of feed it consumes (ipCC, 1997; Garcia-apaza et al., 2008). South africa’s animal data are divided into three main groups according to their different methane-producing abilities, namely ruminants (cattle, sheep, and goats), pseudo-ruminants (horses, donkeys) and monogastric animals (pigs) (DaFF, 2010).

South africa does not have any managed camels or ilamas so these were excluded from the emissions. Buffalo and other game are not managed per say, but are found in significantnumbers ingameparks(bothnationalandprivate). this inventory includes a first estimate of emissions from game in private parks. this number is not complete as not all ruminant species were included due to a lack of emission factor data. Furthermore, an estimate from the game population kept in national parks has not been included due to a lack of population data.

enteric fermentation emissions from poultry were not estimated as the amount produced is considered negligible (ipCC, 2006). No default emission factors are provided in theIPCCGuidelinesasitstatesthatthereisinsufficientdata for the calculation. this exclusion of poultry from enteric fermentation emissions is in line with the ipCC 2006 Guidelines, however, there are some reports of

CH4 emissions from poultry (Wang and Huang, 2005; Burns, 2012). these emissions are small, but in light of South africa’s growing poultry population it should be investigated further in future inventories.


enteric fermentation emissions from livestock declined from 29 290 Gg Co2eq in 2000 to 27 198 Gg Co2eq in 2012 (Figure 5.4: Sector 3 aFolU - livestock: trend and emission levels of enteric fermentation emissions in the livestock categories, 2000 – 2012.), mainly due to a decline in population numbers. emissions declined in most years except for an increase of 2.3% between 2007 and 2008, as well as a 3.0% increase between 2009 and 2010. in these years there was an increase in enteric fermentation emissions from cattle, which was attributed to an increase in the number of mature cows and heifers and a reduction in the calf population (Figure 5.5: Sector 3 aFolU - livestock: trend in cattle herd composition, 2000 – 2012.). Mature cows and heifers have a higher emission factor than calves due to a well-developed gut. Cattle are the largest contributors to the enteric fermentation emissions (78.1% in 2012), with 9.2% from dairy cattle, 13.9% from sheep and 3.2% from goats in 2012 (Figure 5.6: Sector 3 aFolU – livestock: Contribution of the livestock categories to the enteric fermentation emissions in 2012.).


the proportion of intake that is converted into methane is dependent on the characteristics of the animal, feed and the amount eaten. Given the heterogeneity of available feed types within South africa it was considered importanttousemethodologiesthatcouldreflectsuchdifferences, developed under similar conditions as in australia. a detailed description of the methods, data sources and emission factors is found in Du toit et al. (2013a, 2013b), and are summarized below.



Figure 5.5: Sector 3 aFolU - livestock: trend in cattle herd composition, 2000 – 2012.

Figure 5.4: Sector 3 aFolU - livestock: trend and emission levels of enteric fermentation emissions in the livestock categories, 2000 – 2012.


emissions from enteric fermentation are calculated from activity data on animal numbers and the appropriate emission factor:

CH4 emission = ΣeFi (kg CH4 animal-1) * [number of animals for livestock category i]

(eq. 5.1)

5.3.3.3.1Dairycattle[3A1AI]

eFs for the various sub-categories of dairy cattle were taken from Du toit et al. (2013a). these eF’s were calculated using a tier 2 approach following the equation:

eF = ((y/100) * Gei * 365)/ F

(eq. 5.2)

Where:

eF = emission factor in kg CH4 animal-1 year-1;

y = % of Gei yielded as CH4 (calculated from equation in Blaxter and Clapperton, 1965);

Gei = Gross energy intake (MJ head-1 day-1);

F = 55.22 MJ (kg CH4)-1 (Brouwer, 1965).

the gross energy intake (Gei) is the sum of the intake converted into energy terms assuming a gross energy content of 18.4 MJ/kg. an average daily milk production (14.5 kg/day) was sourced from laCto data (2010). Further details of how the various factors were calculated are provided in Du toit et al. (2013a).

Figure 5.6: Sector 3 aFolU – livestock: Contribution of the livestock categories to the enteric fermentation emissions in 2012.



5.3.3.3.2Othercattle[3A1AII]

the eFs for commercial and communal cattle were taken from Du toit et al. (2013a). these eF’s were calculated following an equation developed by Kurihara et al. (1999) to calculate the total daily methane production for animals grazing in tropical pastures:

eF = ((34.9 x i – 30.8)/ 1000) x 365

(eq. 5.3)

Where:

eF = Methane emission factor (kg CH4 animal-1 year-1)

i = Feed intake (kg day-1)

the feed intake (i) was calculated from live weight and live weight gain data following the equation of Minson and McDonald (1987). an additional intake for milk production was incorporated where a feed adjustment value of 1.3 was used during the calving season and 1.1 during the following season. Further details are provided in Du toit et al. (2013a).

Feedlot beef cattle

Feedlot enteric methane emission factors were taken from du toit et al. (2013a) who based their calculations onintakeofspecificdietcomponentsusinganequationdeveloped by Moe and tyrrell (1979):

eF = (3.406 + 0.510Sr + 1.736H + 2.648C) x 365

(eq. 5.4)

Where:

Sr = intake of soluble residue (kg day-1);

H = intake of hemicellulose (kg day-1);

C = intake of cellulose (kg day-1).

Soluble residue intake, hemicellulose intake and cellulose intake were calculated from feedlot diet analysis (aNir, 2009) and average dry matter intake (DMi) taken as 8.5 kg DM per day (South african feedlot association and industry experts).

total annual methane production (eF, kg CH4 head-1 year-1) was calculated as:

eF= (y / F) x 365

(eq. 5.5)

Where:

F = 55.22 MJ (kg CH4)-1 (Brouwer, 1965)

the total feedlot calculations are based on the assumption that an animal will stay in the feedlot for approximately 110 days (3 cycles per year).

5.3.3.3.3 Sheep [3A1C]

Sheep CH4 emission factors were taken from Du toit et al. (2013b) which are based on the equations of Howden et al. (1994):

eF = (i x 0.0188 + 0.00158) x 365

(eq. 5.6)

Where:

eF = emission factor (kg CH4 head-1 year-1);

i = Feed intake (kg DM day-1).

the actual feed intake by sheep was calculated from the potential feed intake (determined by body size and the proportion of the diet that was able to be metabolised by the animal), relative intake (based on dry matter availability) and an additional intake for milk production.


5.3.3.3.4Goats[3A1D]

the CH4 emission factors for the various goat sub-categories were taken from Du toit et al. (2013b). these calculations followed the same methodology as for sheep. Goats are browsers they are also selective feeders and will select for quality. it was assumed that lactating milk goats will receive a higher quality diet with a DMD of 70% throughout the year.

5.3.3.3.5 Swine [3A2H]

CH4 emission factors for the pig sub-categories were taken from Du toit et al. (2013c) which were based on the methodology described in the australian National inventory (aNir, 2009):

eF = i x 18.6 x 0.007 / F

(eq. 5.7)

Where:

eF = emission factor (kg CH4 head-1 year-1)

i = Feed intake (kg DM day-1);

F = 55.22 MJ (kg CH4)-1 (Brouwer, 1965);

18.6 = MJ Ge per kg feed DM.

a methane conversion factor of 0.7% was used in the calculation for pigs, based on the aNir (2009).

5.3.3.3.6 Horses and donkeys [3A1F and 3A1G]

the contribution of horses and donkeys to the total methane emissions was relatively small, and given the lack of data on methane production from these animals, a complex methodology, incorporating relationships between feed intake and methane production, is inappropriate. therefore the emissions from horses and donkeys were based on the tier 1 methodology and ipCC default emission factors.

5.3.3.3.7. Game [3A1J]

eF’s for ruminant game species were taken directly from Du toit et al. (2013d). For hippopotamus and rhinoceros, the eFs were based on the daily methane emissions of elephants (0.1 g CH4/kg lW/day) as was done in Du toit et al. (2013).


5.3.3.4.1Livestockcategorizationandpopulationnumbers

the f irst step in estimating CH4 emissions from enteric fermentation and CH4 and N2o emissions from manure management involves dividing the farmed animal population into the livestock categories and sub-categories. the livestock categories and sub-categories used in this inventory update are shown in Table 5.2. the sub-categories for sheep, goats and swine were obtained by combining some of the more detailed sub-categories given in Du toit et al. (2013a, 2013b, 2013c) (see Appendix E) as some of the populations were small andEF’sdidn’tvarysignificantly.Theinclusionofallthedetailed sub-categories should be considered in future inventories.

the commercial cattle (dairy and other), sheep, goat and swine population data were obtained from the abstracts of agricultural Statistics 2012. the population numbers for these animals has a consistent time-series dating back to 1970. For other cattle the feedlot cattle need to be separated out as feedlot cattle have a different emission factor. annual feedlot cattle numbers were obtained from the South african Feedlot association (Sa Feedlot association, 2013; Ford, Dave, 2013. pers. Comm.). Data was only available for the years 2008 to 2012, so an average of 420 000 was used for each year between 2000 and 2007. the number of cattle in feedlots fluctuatesbyatmost20000aroundthisnumberandthisnumber is not expected to change much over a 10 year period (Ford, Dave, 2013). the feedlot population was assumed to consist of calves, heifers and young oxen in



tabl

e 5.

2:

Sect

or 3

aFo

lU –

liv

esto

ck: l

ives

tock

cat

egor

ies

used

in t

he d

eter

min

atio

n of

live

stoc

k em

issi

ons.

Mai

n ca

tego

ry

Dai

ry c

attl

eO

ther

ca

ttle

Shee

pG

oats

Swin

eH

orse

sM

ules

an

d as

ses

Gam

eT

MR

a -ba

sed

Past

ure-

base

d

Sub-category

Commercial

lact

atin

g co

wla

ctat

ing

cow

Bulls

Non

-woo

l ew

eBu

ckBo

ars

elep

hant

lact

atin

g H

eife

rla

ctat

ing

Hei

fer

Cow

sN

on-w

ool r

amD

oeSo

ws

Gir

affe

Dry

cow

Dry

cow

Hei

fers

Non

-woo

l lam

bK

ids

pigl

ets

elan

d

preg

nant

hei

fers

preg

nant

hei

fers

ox

Woo

l ew

ea

ngor

aBa

cone

rs

Bu

ffalo

Hei

fers

>1

yrH

eife

rs >

1 yr

youn

g ox

Woo

l ram

Milk

goa

tpo

rker

s

Z

ebra

Hei

fers

6-1

2 m

ths

Hei

fers

6-1

2 m

ths

Cal

ves

Woo

l lam

b

K

udu

Hei

fers

2-6

mon

ths

Hei

fers

2-6

mon

ths

Feed

lot

Wat

erbu

ck

Cal

ves

< 2

mon

ths

Cal

ves

< 2

mon

ths

Blue

Wild

e-be

est

Blac

k W

ilde-

bees

t

tse

sseb

e

Bles

bok

War

thog

impa

la

Spri

ngbo

k

Hip

popo

tam

us

rhi

noce

ros

Communal

Bulls

Non

-woo

l ew

eBu

ckBo

ars

Cow

sN

on-w

ool r

amD

oeSo

ws

Hei

fers

Non

-woo

l lam

bK

ids

pigl

ets

ox

Woo

l ew

eBa

cone

rs

youn

g ox

Woo

l ram

pork

ers

Cal

ves

Woo

l lam

b

a tot

al m

ixed

rat

ion


an equal split. therefore the total commercial calf, young oxen and heifer population numbers from the abstracts of agricultural Statistics (DaFF, 2012) were reduced by a third of the feedlot population for that year and then the feedlot category was added into the population.

the herd composition data from Du toit et al. (2013a, 2013b, 2013c) was applied to the national dairy, sheep and goat population numbers to obtain the herd composition sub-categories for these livestock. the average annual herd composition for swine was assumed to be 1.9% boars, 32.2% sows, 33% piglets, 4.9% porkers and 28% baconers (Du toit et al., 2013c).

Due to a lack of data over time on herd composition it was assumed that the composition remained constant between 2000 and 2010. in future this assumption needs to be addressed by improving the monitoring of livestock herd composition. altering the composition of the herd is a possible, although small, mitigation action therefore improved monitoring of herd composition is required in the future to address this assumption.

the total communal cattle population was determined from the difference in the total cattle population data provided in Table 58 of the abstracts of agricultural Statistics (2013) and the total commercial dairy and other cattle provided in Table 59 of the abstracts. all the communal cattle were assumed to be other cattle. the communal population was assumed to have the same herd composition (excluding feedlot cattle) as the commercial population. the total communal population numbers for sheep, goats and swine was obtained by using the ratio of commercial to communal population from the quarterly census numbers which have been recorded by DaFF from 1996 onwards. the ratios were 0.1396 for sheep, 1.9752 for goats and 0.1306 for swine. the communal livestock numbers were given as a total so to determine the number in each livestock sub-category the same herd composition as that of commercial livestock was assumed.

Data on horse and donkey populations in South africa are very scarce and the data that is available varies greatly.

Census data obtained from DaFF showed there were 20 588 horses in 2002 and 21 431 in 2007. However, a detailed report by Simalenga et al. (2002) indicated the horse population to be 180 000, while the donkey population was at 1 million. this was however based on 1995 data. Fao data (http://faostat.fao.org/site/573/default.aspx#ancor; accessed on 14 april 2014) gives a constant value of 270 000 for horses and 164 000 mules/asses in South africa between 2000 and 2005. after which the numbers increase slightly. the population data varies quite widely, and it was decided to use the more consistent Fao data set for horses, mules and asses. population data for all game animals was taken from Du toit et al. (2013d). all livestock population numbers are provided in Appendix E.

5.3.3.4.2 Live weight data

the live weight (or typical animal mass [taM]) data of the various livestock categories and sub-categories was taken from Du toit et al. (2013a, 2013b, 2013c) (Table 5.3). For sheep, goats and swine where some of the subcategories were combined a weighted average taM (based on population numbers) was calculated for the livestock subcategories used in this inventory. For game the live weight data was taken from Du toit et al. (2013d) except for hippo and rhino. the live weight for hippo and rhino were taken to be the average weight provided in Clauss et al. (2003). For rhino it was the average weight of both black and white rhino.

5.3.3.4.3 Methane emission factors

SouthAfricahasidentifiedentericfermentationasakeysource category; therefore, tier 2 methods were used to determine enteric fermentation emissions from the major livestock categories, as well for game. the eFs were taken from Du toit et al. (2013a, 2013b, 2013c, 2013d) (Table 5.3 and Table 5.4), who developed country-specificmethodologiesbasedonthemethodsdeveloped by australia and described in detail in their National inventory reports (aNir, 2009). in the 2004



agricultural inventory (DaFF, 2010) it was determined that the eFs for South african livestock were similar to those of australia, a country which has similar climatic conditions. this was one of the reasons for adopting the australian methodology. the eF takes into account the climate, feed digestibility and energy intake of the various livestock. in some cases there were more detailed

subcategories, so in these cases a weighted average eF was calculated for the livestock subcategories used in this inventory. Du toit et al. (2013a, 2013b, 2013c, 2013d) calculated the emission factors for the year 2010. in this inventory, we assumed that the eFs remained constant between 2000 and 2010. ipCC 2006 default eFs were used for horses, mules and asses.

table 5.3: Sector 3 aFolU – livestock: livestock weights, methane eF’s and ipCC 2006 default eF’s.

Livestock category Livestock sub-categoryTAM (kg)

Calculated EFenteric

IPCC default EF (kg/head/year)

(kg/head/year) africa oceania

Dairy – tMr

lactating cows 590 132.00

40 81

lactating heifers 503 127.00Dry cows 590 80.41pregnant heifers 394 67.66Heifers > 1 year 322 62.63Heifers 6 – 12 months 172 42.12Heifers 2 – 6 months 55 22.50Calves 35 21.51

Dairy – pasture

lactating cows 540 127.00

40 81

lactating heifers 438 116.00Dry cows 540 83.37pregnant heifers 333 61.78Heifers > 1 year 254 52.63Heifers 6 – 12 months 142 37.11Heifers 2 – 6 months 54 24.49Calves 36 20.02

other cattle – commercial

Bulls 733 112.63

31 60

Cows 475 92.06Heifers 365 75.89ox 430 89.44young ox 193 51.64Calves 190 51.58Feedlot 335 58.87

other cattle – communal

Bulls 462 83.83

31 60

Cows 360 73.09Heifers 292 62.51ox 344 72.56young ox 154 41.58Calves 152 40.92


Livestock category Livestock sub-categoryTAM (kg)

Calculated EFenteric

IPCC default EF (kg/head/year)

(kg/head/year) africa oceania

Sheep – commercial

Non-wool ewe 60 9.07

5 8

Non-wool ram 78 11.47Non-wool lamb 29 4.39Wool ewe 55 8.27Wool ram 92 13.83Wool lamb 28 4.31

Sheep – communal

Non-wool ewe 48 6.465 8Non-wool ram 62 8.13

Non-wool lamb 23 3.30Wool ewe 44 5.95

5 8Wool ram 70 9.80Wool lamb 22 3.24

Goats – commercial

Buck 106 16.25

5 5Doe 74 11.30Kid 32 4.82angora 24 3.91Milk goat 39 6.49

Goats – communalBuck 68 9.54

5 5Doe 51 6.93Kid 21 3.05

Swine – commercial

Boars 253 2.06

1.0 1.5Sows 306 2.29piglets 9 0.43Baconers 90 0.99porkers 70 0.51

Swine – communal

Boars 212 1.65

1.0 1.5Sows 245 1.84piglets 7 0.34Baconers 70 0.79porkers 56 0.41

Horses 595 18 18 18

Mules and asses 250 10 10 10



table 5.4: Sector 3 aFolU – livestock: Weights and methane eF’s for game animals.

Livestock category Livestock sub-category TAM (kg)EFenteric

(kg/head/year)

Game

elephant 2436 81.0

Giraffe 826 136

eland 528 93.7

Buffalo 466 113

Zebra 266 13.9

Kudu 155 31.3

Waterbuck 150 35.9

Blue wildebeest 153 24.8

Black wildebeest 106 14.3

tsessebe 105 13.8

Blesbok 62 9.1

Warthog 59 2.2

impala 42 7.4

Springbok 28 4.7

Hippopotamus 1453 47.5

rhinoceros 5452 62.2

Comparison of methane EF with IPCC defaults

a comparison of the calculated methane eFs for cattle, sheep, goats and swine and the ipCC default factors is provided in Table 5.5. the eFs were seen to be in the same range as the oceania or developed-country eFs, as opposed to the africa or developing-country default factors. this was not unexpected. the reasons for these differences were evaluated by investigating the ipCC 2006 default productivity data used to calculate the default emission factors. the milk production in South africa in 2010 was 14.5 kg day-1, which is much higher than the 1.3 kg day-1 given for africa (Table 5.5). the cattle weights in South africa were much higher than those given for the african default. For example, the weight of dairy cows in this study was between 333 kg and 590 kg, which were much higher than the 275 kg given for

africa. the pregnancy and De percentages in this study were also higher than those used in the calculation of the default ipCC values.


Uncertainty in livestock population numbers is around 10% (Meissner, pers. Comm., 2015) except for dairy and swine population data which is slightly higher (15%). emission factor uncertainty has not been calculated but the ipCC suggests the uncertainty for tier 2 eF is around 20%. overall uncertainty for enteric fermentation for the various livestock was therefore determined to be between 9% and 20%. time-series consistency was ensured by using consistent methods for the 12 year period, as well as for the recalculations in 2000.



population data was obtained for a 40-year period for cattle, sheep, goats and swine, and trends were checked. there were no sudden changes in the data. For poultry, a 12-year trend was monitored. population numbers were also checked against Fao data and numbers from the individual livestock organizations for the year 2012. these numbers were found to vary slightly, however, the variation for all livestock was smaller than the uncertainty on the national population numbers, and therefore the national numbers were used because of the long time period of the data. it also assists with the time-series consistency if the same source of data is used throughout. Swine population data in this inventory and in Fao data aremuchhigherthanthefiguresreportedbyDuToitet al. (2013c), so the reason for the discrepancy should be investigated in future.

emission factors were compared to the ipCC default factors. No actual measurements have been made on methane emissions from livestock in South africa, so no direct comparisons can be made.



the other cattle population data was recalculated for all years between 2000 and 2012, as it was determined that feedlot cattle were already incorporated into the commercial population data. also, the ratio of communal to commercial cattle was adjusted based on data from the abstracts of agricultural Statistics (2013). in this

table 5.5: Sector 3 aFolU – livestock: Comparison between the productivity data used and eF calculated in this inventory and the ipCC 2006 Guideline default values.

Livestock category

ParameterValue used in this

study (2010)

IPCC 2006 default values(table 10.11, 10a.1, 10a.2)

africa oceaniaWestern europe

Dairy – tMr

Milk production 6,015 475 2,200 6,000

(kg head-1 yr-1)Milk production

14.5 1.3 6 16.4(kg day-1)

Cattle weight (kg)322 – 590 275

500 600(heifer - mature cow) (mature cow)

pregnancy (%) 58 67 80 90De% 76 60 60 70

other cattle

lactating cows

Communal: 200 – 275 400 – 450 400 – 600360 – 462 (cow - bull); (cow - bull) (cow - bull) (bulls)

Commercial:

475 – 733 (cow - bull)

pregnancy 24 – 49 33 67No value

givenDe% 55 – 80 55 55 60 – 65



inventory, emissions from ruminant game in privately owned game parks were estimated. there were not in the previous inventory. Due to limited population data the numbers and emissions for game were kept constant between 2000 and 2012, so a constant emission from game was added to all years.


there are no planned improvements, however it is recommendedthatifsufficientdataisavailableannualemission factors, incorporating changes in feed quality and milk production, should be considered instead of assuming a constant emission factor across all years. it is also recommended that a more thorough uncertainty analysis be completed, following the ipCC 2006 Uncertainty Guidelines and Good Practice Guidelines.

5.3.4 Manure management [3A2]

5.3.4.1 Source-Category Description

Manure management includes the storage and treatment of manure, before using it as farm manure or burning it as fuel. Methane (CH4) and nitrous oxide (N2o) are produced during different storage and treatment stages of manure. the term “manure” includes both dung and urine produced by livestock.

Camels and ilamas do not occur in South africa and were therefore not included in the report. there is a large population of game in South africa, but a lack of data hinders the inclusion of their emissions. National population data for ruminant game is lacking, however, there is population data for game kept in privately owned parks. Du toit et al. (2013) determined that manure CH4 emissions from game were very low and therefore no emission factors were reported. this inventory thus assumes zero manure CH4 emissions from game.

the majority of manure in South africa is produced and left in the pastures while grazing, with only a small amount of manure is being managed in South africa. N2o manure emissions are reported for managed manure on dairy, swine, feedlot and poultry. in accordance with ipCC Guidelines, N2o emissions from the manure left in pastures and daily spread are not taken into account in this source category (manure management), but are included in the source category for managed soils.


Manure management produced a total of 2 214 Gg Co2eq in 2012, which was an increase from the 1 872 Gg Co2eq produced in 2000 (Figure 5.7: Sector 3 aFolU – livestock: total manure management trend and emission levels from source categories, 2000 – 2012.).

Manure emissions showed the greatest increase between 2004 and 2007, with annual increases of 4.0%, 2.6% and 2.8% during this period. there was also another large increase between 2010 and 2011. emissions generally increased annually, with the only decline (0.5%) occurring between 2002 and 2003. these increases are due to increased N2o emissions from poultry manure, but N2o emission from feedlot manure also contributed to the increase between 2010 and 2011. this was mostly related to increases in population numbers during that time.

Methane emissions accounted for 38.5% of manure emissions in 2012, with its contribution slowly declining from 46.2% in 2000. Manure CH4 emissions showed peaks in 2001 (875 Gg Co2eq) and in 2009 (870 Gg Co2eq) (Figure 5.8: Sector 3 aFolU – livestock: Manure management CH4 trend and emission levels from source categories, 2000 – 2012.). the largest contributor to the CH4 emissions were swine (68.9% in 2012), followed by dairy cattle (18.1% in 2012). the contribution of CH4 from dairy cattle manure management decreased by 1.6% since 2000, while the contribution from swine manure declined by 1.9% over the same period.


Figure 5.8: Sector 3 aFolU – livestock: Manure management CH4 trend and emission levels from source categories, 2000 – 2012.

Figure 5.7: Sector 3 aFolU – livestock: total manure management trend and emission levels from source categories, 2000 – 2012.



N2o dominated the manure emissions, contributing 57.5% in 2012. the contribution has slowly increased since 2000 (when it contributed 53.8%). N2o emissions from manure management increased from 1 007 Gg Co2eq (2000) to 1 362 Gg Co2eq (2012), mainly due to the increased manure from poultry. poultry contribution has increased from 63.8% in 2000 to 70.1% in 2012, while the contribution from cattle feedlots declined by 4.4% over the same period (Figure 5.9: Sector 3 aFolU – livestock: Manure management N2o trend an emission levels from source categories, 2000 – 2012.). Swine manure N2o emissions declined by 4.1% between 2000 and 2012.

the total emissions from poultry manure increased by 48.6% between 2000 and 2012, and their contribution to manure emissions increased from 36.2% to 45.7%.


the methane produced from manure management of dairy cattle, beef cattle and feedlot cattle was calculated based on the approach of the ipCC (2006), using a combination ofdefaultIPCCandcountry-specificinputvalues.Duetothe similarity in environmental and climatic conditions, the australian methodology was used to calculate the manure emission factor (MeF) of range-kept cattle in environments with an average temperature of 21oC. this was determined to be 1.4 x 10-5 kg CH4 (kg DM manure)-1 based on Gonzalez-avalos and ruiz-Suarez (2001).

Figure 5.9: Sector 3 aFolU – livestock: Manure management N2o trend an emission levels from source categories, 2000 – 2012.


5.3.4.3.1 CH4 emissions from animal manure

Dairy cattle [3A2ai]

Methane emissions from manure originate from the organic fraction of the manure (volatile solids). Manure CH4 emissions were taken from Du toit et al. (2013a). a tier 2 approach was used and the emissions (kg head-1 year-1) were determined using the equation:

eF = (vS x Bo x MCF x p ) x 365

(eq. 5.8)

Where:

Bo = emissions potential (0.24 m3 CH4 (kg vS)-1; ipCC, 2006);

MCF = integrated methane conversion factor – based on the proportion of the different manure management systems;

p = density of methane (0.662 kg m-3)

Using the methane conversion factor (MCF) from the ipCC 2006 Guidelines for the various manure management systems the integrated MCF for lactating dairy cattle in total mixed ration (tMr) based production systems was calculated as 10% and 1% for all other classes of dairy cattle. in pasture based production systems the integrated MCF for lactating cattle was calculated as 4.57% and 1% for all other classes of cattle. volatile solids (vS, kg head-1 day-1) were calculated according to aNir (2009) as:

vS = i x (1 – DMD) x (1 – a)

(eq. 5.9)

Where:

i = dry matter intake calculated as described above;

DMD = Dry matter digestibility expressed as a fraction;

a = ash content of manure expressed as a fraction (assumed to be 8% of faecal DM, ipCC 2006 Guidelines).

Other cattle, sheep and goats [3A2aii, 3A2c and 3A2d]

Methane emissions from manure (kg head-1 year-1) of all categories of beef cattle, sheep and goats were taken from Du toit et al. (2013a and 2013b) who followed the methods used in the australian inventory (aNir, 2009):

eF = (i x (1 – DMD) x MeF) x 365

(eq. 5.10)

Where:

i = intake as calculated under enteric emissions (Section 5.3.3.3);

MeF = Manure emissions factor (kg CH4 (kg DM manure)-1).

Feedlot cattle

the ipCC default methane conversion factor for drylot (1.5%) was used. the volatile solid production for feedlot cattle was estimated based on data developed under the enteric methane emission calculations. the volatile solid production and methane emission factor was calculated as for dairy cattle (eq. 5.9), but assuming a DMD of 80% for feedlot diets and a B0 of 0.17 m3 CH4 (kg vS)-1 (ipCC, 2006).

Swine [3A2h]

Commercial pig production systems in South africa are housed systems and a large proportion of manure and waste is managed in lagoon systems. the CH4 emission factors for pig manure (kg head-1 year-1) were taken from Du toit et al. (2013c) and calculated following the



equations provided in aNir (2009):

eF = (vS x Bo x MCF x p ) x 365

(eq. 5.11)

Where:

vS = volatile solid production (kg head-1 year-1);

Bo = emissions potential (0.45 m3 CH4 (kg vS)-1) (ipCC 2006);

MCF = integrated methane conversion factor calculated from ipCC 2006 MCF for each manure management system;

p = density of methane (0.622 kg m-3).

Du toit calculated the volatile solid production from South african pigs by using the following equation provided in the ipCC 2006 Guidelines:

vS = [Ge x (1 – (De%/100)) + (Ue x Ge)] x [(1 – aSH)/18.45]

(eq. 5.12)

Where:

vS = volatile solid excretion (kg vS day-1);

Ge = Gross energy intake (MJ day-1);

De% = digestibility of feed (%);

(Ue x Ge) = urinary energy expressed as a fraction of Ge. typically 0.02Ge for pigs (ipCC 2006);

aSH = ash content of manure (17%, Siebrits, 2012);

18.45 = conversion factor for dietary Ge per kg of DM (MJ kg-1).

Poultry [3A2i]

as for the other manure CH4 emission factors, the poultry emission factors were calculated according to equation 5.8, using an integrated MCF of 1.5% (ipCC, 2006; Du toit et al., 2013c). volatile solid production from poultry production systems was calculated using the same equation as for dairy cattle (eq. 5.9), but assuming a dry matter intake of 0.11 kg day-1, a DMD of 80% and an ash content of 8% of faecal DM (aNir, 2009).

Game [Other - 3A2j]

CH4 emissions from game manure were assumed to be zero as Du toit et al. (2013d) indicated that emissions were minimal and did not report any eFs.

5.3.4.3.2 N2O emissions from animal manure

Du toit et al. (2013a) calculated nitrogen excretion from the various livestock categories using crude protein inputs and storage, and from this the output of nitrogen in faeces and urine (following the methodology provided in aNir, 2009).Unfortunately, therewere insufficientdatatoextend these calculations from 2010 to the period 2000 to 2009, therefore a tier 1 approach was used, as in the 2004 agricultural inventory (DaFF, 2010). N2o emissions from manure management are not a key source of emissions sotheTier1methodologyissufficient.

Direct N2o emissions from manure management were calculated from animal population data, activity data and manure management system data using equation 10.25 and 10.30 from the ipCC 2006 Guidelines.


population data sources for all livestock except poultry are described in Section 5.3.3.4.1 and the typical animal masses (taMs) are provided in Table 5.3: Sector 3 aFolU – livestock: livestock weights, methane eF’s and ipCC 2006 default eF’s. poultry were also included in


the manure management emission calculations. the total number of layers and slaughtered broilers was obtained from the South african poultry association (prinsloo, Magda, 2014, pers. Comm.). the layer population numbers were used as they were provided, but, for broilers, the annual average population (aap) was calculated, as a broiler is not alive for 365 days a year. the aap was determined using the average number of broiler growing cycles (per annum), using an equation similar to that used by New Zealand (Ministry of agriculture and Forestry, 2011):

aap = Napa/r (eq. 5.13)

Napa is the number of animals produced annually, which was taken to be the number of broilers slaughtered; and r is the average number of broiler growing cycles per annum (r = 8, Du toit et al., 2013). it has been noted that New Zealand changed their methodology from using the growing cycle to using the number of days alive. it indicated that the former method was leading to an overestimation. However, for South africa, using the growing cycle brings the population numbers more in line with data from the census that was carried out between 2000 and 2013. Communal layers and broilers were calculated as 4.2% of the commercial population (prinsloo, Magda, 2014, pers. Comm.). the taM for poultry was 1.8kg for commercial broilers, 2 kg for layers and 0.9 kg for communal broilers (DaFF, 2010; ipCC, 2006). population data is provided in Appendix D.

5.3.4.4.1 Manure management systems

For dairy cattle it is only the manure from lactating cows and heifers that is managed (Table 5.6), as the rest of the herd is kept in the pasture so manure is deposited in the pasture, range and paddock (Du toit et al., 2013a). all other cattle, sheep, goats, horses and donkeys are considered range-fed and therefore all manure gets depositedinthefieldandisnotmanaged(DuToitet al., 2013a, 2014b, DaFF, 2010). all swine manure is managed

either as lagoon, liquid/slurry, drylot or daily spread (Du toit et al., 2013c; Table 5.6). all cattle feedlot manure and poultry manure is managed as drylot (Du toit et al., 2013a; DaFF, 2010).

5.3.4.4.2 CH4 activity data and emission factors

volatile solids (vS) for cattle, sheep, goats, pigs and poultry were calculated using methods from the aNir (2009) and are discussed in Section 5.2.2.4 below. the africa default vS values were used for horses and donkeys (ipCC, 2006, tables 10a.4-10a.9). Methane conversion factors (MCF) were obtained from the ipCC 2006 Guidelines. all other activity data were obtained from Du toit et al. (2013a, 2013b).

emission factors for cattle, sheep, goats, swine and poultry (Table 5.7) were taken from Du toit et al. (2013a, 2013b, 2013c) and the methods are described in Section 5.3.3.3. in the case of sheep, goats and swine there were additional livestock subcategories in Du toit et al. (2013a, 2013b, 2013c) (see Appendix E) and in these cases a weighted average emission factor was calculated. ipCC 2006 africa default values were used for horses and donkeys.

Comparison of manure EF’s and IPCC defaults

a comparison of the calculated manure emission factors with ipCC default values is given in Table 5.7. For dairy cattle the eF for lactating cows and heifers was much higher than the africa default, while the rest of the dairy cattle were only slightly higher. if an average dairy cattle eF was calculated from this data the eF would be close to the australian eF. For other cattle the eF are lower than the africa default values, but are in a similar range to the australian eF. all cattle eF are lower than the oceania ipCC 2006 default values. Swine eF were much higher than the africa default values and are in a similar range to those of oceania and australia. poultry manure eF for N2o was lower than all default values.



5.3.4.4.3 N2O activity data and Emission factors

Nitrogen excretion rates (Nrate) were obtained from the africa default values (ipCC, 2006, table 10.19) while the annual N excretion for livestock Nex was estimated using equation 10.30 from the guidelines (ipCC, 2006). the typical animal mass (taM) for the various livestock categories is given in Table 5.3. the manure management system usage data (Table 5.6) was used to produce the fraction of total annual nitrogen excretion for each livestock category managed in the various manure management systems.

table 5.6: Sector 3 aFolU – livestock: Manure management system usage (%) for different livestock categories, 2000 – 2012 (Source: DaFF, 2010; Du toit et al., 2013a and 2013b).

Livestock Category

Sub-category LagoonLiquid /slurry

DrylotDaily

spreadCompost Pasture

Country specificclinker fraction

tMr - lactating cows/heifers

10 0.5 0 1 0 88.5

pasture - lactating cows/heifers

3 0 0 7 0 90

all other 0 0 0 0 0 100

other Cattle

Feedlot cattle 0 0 100 0 0 0

Commercial 0 0 0 0 0 100

Communal 0 0 0 0 0 100

SheepCommercial 0 0 0 0 0 100

Communal 0 0 0 0 0 100

GoatsCommercial 0 0 0 0 0 100

Subsistence 0 0 0 0 0 100

Horses 0 0 0 0 0 100

Donkeys 0 0 0 0 0 100

pigsCommercial 92 1.5 5 1.5 0 0

Communal 0 0 50 50 0 0

poultrylayers 0 0 100 0 0 0

Broilers 0 0 100 0 0 0

ipCC 2006 default N2o emission factors were used for the various manure management systems (ipCC 2006, table 10.21).


table 5.7: Sector 3 aFolU – livestock: CH4 manure eF’s for the various livestock compared with the ipCC 2006 default factors and the australian eF’s (Source: Du toit et al., 2013a, b, c; ipCC 2006 Guidelines; aNir, 2009).

Livestock Category

Sub-category

Lagoon Liquid /slurry

Drylot Daily spread

(kg/h/year) africa oceania (kg/head/year)

Dairy

-

TMR

lactating cows 14.800

1 29 8.8

lactating heifers 14.700

Dry cows 1.470

pregnant heifers 1.240

Heifers > 1 year 1.190

Heifers 6 – 12 months 0.750


Calves 0.210

Dairy

-

pasture

lactating cows 4.980

1 29 8.8

lactating heifers 4.800

Dry cows 1.110

pregnant heifers 0.880

Heifers > 1 year 0.780



Calves 0.320

Other cattle

-

commercial

Bulls 0.022

1 20.04

Cows 0.018

Heifers 0.016

ox 0.018

young ox 0.012

Calves 0.012

Feedlot 0.870 2.91

other cattle

-

communal

Bulls 0.017

1 2 0.04

Cows 0.015

Heifers 0.013

ox 0.015

young ox 0.010

Calves 0.010

Sheep

-

commercial

Non-wool ewe 0.003

0.15 0.28 0.002

Non-wool ram 0.003

Non-wool lamb 0.001

Wool ewe 0.002

Wool ram 0.004

Wool lamb 0.001



Livestock Category

Sub-category

Lagoon Liquid /slurry

Drylot Daily spread

(kg/h/year) africa oceania (kg/head/year)

Sheep

-

communal

Non-wool ewe 0.002

0.15 0.28 0.002

Non-wool ram 0.003

Non-wool lamb 0.001

Wool ewe 0.002

Wool ram 0.003

Wool lamb 0.001

Goats

-

commercial

Buck 0.018

0.17 0.2

Doe 0.012

Kid 0.005

angora 0.004

Milk goat 0.005

Goats

-

communal

Buck 0.011

0.17 0.2Doe 0.008

Kid 0.004

Swine

-

commercial

Boars 17.470

1 13 - 24 23

Sows 24.190

piglets 3.740

Baconers 20.960

porkers 17.960

Swine

-

communal

Boars 0.390

1 13 - 24 23

Sows 0.540

pre-wean piglets 0.080

Baconers 0.460

porkers 0.400

Horses 1.640 1.64 2.34

Mules and asses 0.900 0.9 1.1

Poultrylayers 0.000 0.03

0.04Broilers 0.000 0.02



5.3.4.5.1 Uncertainty

Data on manure management storage systems under different livestock categories are lacking, with estimates being based on expert opinion and information from the various livestock industries. the uncertainty on the default Bo estimates is ±15%. For volatile solids the uncertainty is ± 20% for dairy cattle, ± 35% for other cattle and ± 25% for pigs. Country average temperatures were used and this leads to inaccuracies in the estimates. the uncertainty on the CH4 eF is ± 20% (ipCC 2006 Guidelines). the uncertainty on the default eFs for horses, mules and asses is ± 30%.

the default Nrate values were used for africa and these had an uncertainty of ±50%. Uncertainty on management system usage data is estimated to be between 25% and 50% (ipCC 2006 Guidelines). the drylot eF has an uncertainty of a factor of 2.

overall uncertainties for the various livestock for manure CH4 were determined to be between 12% and 20%, whereas for manure N2o uncertainty is around 220% due to the large uncertainties on eF’s.

5.3.4.5.2 Time-series consistency

time-series consistency was ensured by the use of consistent methods and full recalculations in the event ofanymethodrefinement.


Methane emission factors were compared to ipCC default values (Table 5.7), as well as those used in the australian inventory. No direct measurements of methane emissions from manure have been made in South africa so the calculatedvaluescouldnotbeverified,onlycomparedto default values. Default values were used to determine the N2o emissions from manure management so no data comparisons were made.



recalculations were completed for all years between 2000 and 2012 with the updated population data for other cattle and poultry. poultry data were adjusted for all years based on new information and to bring the numbers more in line with the annual census data. Game manure emissions were considered for all years, but the emissions were assumed to be zero, so no changes were made for this.

5.3.4.8 Source-specificPlannedImprovementsandRecommendations

Nospecificimprovementsareplanned,however,it issuggested that in order to improve the accuracy and reduce the uncertainty of the manure management emission data it would be important to improve the monitoring of manure management systems. another improvement would be to stratify the livestock population by climate, or even provincial data, so that a more accurate weighted average emission factor can be determined.

N2o emission data from manure management systems would also be improved if N excretion rates for cattle in South africa were determined so that actual data could be used instead of the value calculated using ipCC 2006 default values.

5.4 Land [3b]

5.4.1 Overview of the sub-sector

the land component of the aFolU sector includes Co2

emissions and sinks of the carbon pools above-ground



and below-ground biomass, and soils from the categories forest land (3.B.1), croplands (3.B.2), grasslands (3.B.3), wetlands (3.B.4), settlements (3.B.5), other lands (3.B.6), and the relevant land-use change categories. the N2o emissions from the land sector were estimated in the aggregated and non-CO2 emission sources on land section and CH4 emissions from wetlands were also included, following the methodology in the previous inventories (Department of environmental affairs and tourism, 2009 & Department of environmental affairs, 2014). all other emissions in the land category were assumed to be negligible.

according to the maps developed in section 5.4.4 (Geoterraimage, 2015) the land cover in South africa in 2014 was dominated by low shrubland (33.5%) and grasslands (20.7%). Natural forests are very small in South africa covering less than 0.5% of the land area, while settlements occupy approximately 2.3%. agricultural activities cover about 11.2% of the land area, with maize and wheat being the dominant annual crops by area. perennial crops (orchards and viticulture) contribute less than 0.5% towards the cropland land area. plantations are based on non-native trees, the dominant species being Eucalyptus grandis, and they cover about 1.5% of the land area (Forestry South africa, 2012).

Classifying the South african soils into the six main types provided by the ipCC shows that the high-activity clay mineral soil dominates (>60%) (Moeletsi et al., 2013). this soil type consists of lightly to moderately weathered soils, dominated by 2:1 silicate clay minerals, including vertisols, mollisols, calcareous soils, shallow soils and various others. the other two main soil types are sandy mineral soil and low-activity clay mineral soil. Sandy minerals predominate over the Northern Cape, North West and Western Cape. the low-activity clay soils are found mainly in the warmer, higher rainfall areas, such

as KwaZulu-Natal and Mpumalanga. organic soils are considered negligible.

long-term (1920 – 1999) climate data for South africa were used to categorize the climate into the ipCC climate classes. over 95% of South africa is categorised under the “warm temperate dry” climate class. there are a few very small patches over the limpopo ranges, along the KwaZulu-Natal coast, in some parts of Mpumalanga and over the Western Cape, which fall into the “warm temperate moist” class. the other exception is the “cool temperate dry” regions in the ranges of the eastern Cape that are close to lesotho.

the inventory makes use of available data in this sector, however there are still large data gaps. the inventory wouldbenefitfromadetailedgapanalysiswhichincludesnot only a clear description of current data and data gaps, but also a detailed proposals for how these data gaps maybefilledinthefuture(potentiallywithreferenceto organisations collecting data, surveys that could be expanded to collect additional data, etc.).

5.4.2 Overview of shares and trends

5.4.2.1 Land Cover Trends

land-cover change was determined as discussed in Section 5.4.4 and results show that between 1990 and 2014 the Forest land (which includes woodlands and open bush) and settlements increased by 15.4% and 6.7% respectively. Wetlands decreased by 17.6%, while grasslands and croplands declined by 6.2% and 1.6% respectively (Figure 5.10).2 these numbers, however, should be considered in light of the limitations discussed in Section 5.4.4.4.

2 Acautionarynote:therearethreebasicfactorsthatmayhaveinfluencedland-coverchangeresultsandassociatedstatistics,namelyland-cover mapping accuracies, seasonal / climatic conditions under which the original source imagery was acquired, and the level of (landscape) detailthatisbeingcomparedovertime.TheseinfluencingfactorsarediscussedindetailinAppendix G (section 14.2).


5.4.2.2 Trends in CO2

the land sector was estimated to be a net sink of Co2, butthissinkfluctuatedoverthe12yearperiod(Figure 5.11). the biomass and soil carbon pools contribute the most to the land sector (Figure 5.12), however the DoM pool is incomplete

Forest land (3B1) comprised emissions and removals from the above and below ground, soil carbon and DoM pools for the forest land remaining forest land and land converted to forest land sub-categories. Forest lands include plantations, natural forests, thickets, and woodland/open bush. Carbon stock changes due to wood harvesting, fuelwoodconsumption,anddisturbance(includingfires)were all included. the forest land category was estimated to be a net sink for Co2 in all years between 2000 and

2012, varying between a low of 18 988 Gg Co2 (2008) and a high of 34 325 Gg Co2. the land converted to forest land subcategory accounted for 85.9% of the forest land Co2 in 2012 and this is mainly due to the land conversions accumulating in this category until the 23 year period is reached. Furthermore, due to a lack of data the carbon losses could not be separated between land remaining land and converted land. therefore all losses were allocated to land remaining land, leading to the converted lands having a higher sink.

Croplands (3B2) comprised emissions and removals from the biomass and soil carbon pools in the croplands remaining croplands and the land converted to croplands sub-categories. Croplands include annual crops, perennial crops (viticulture, orchards), and cultivated subsistence crops. Cropland had an emission of 4 480 Gg Co2

Figure 5.10: Sector 3 aFolU - land: Comparison of land areas and the change between 1990 and 2013-2014.



Figure 5.12: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the land sub-sector between 2000 and 2012, differentiated by source category.

Figure 5.11: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the land sub-sector between 2000 and 2012, differentiated by sub-category.


in 2000 and this increased to 5 616 Gg Co2 in 2012. Croplands remaining croplands are generally assumed to be in balance, however there were some conversions between the different crop types which lead to a sink in this category which declined to 104 Gg Co2 from 1 242 Gg Co2 in 2000.

Grasslands (3B3) comprised emissions and removals from the biomass and soil carbon pools for grasslands remaining grasslands and land converted to grasslands subcategories. the grassland category only includes grassland vegetation. the grassland remaining grasslands subcategory, following the tier 1 methodology, are assumed to be in balance, while land converted to grasslands were a source of 7 904 Gg Co2 annually.

Wetlands (3B4) were estimated not to have any emissions or removals of Co2, but CH4 emissions were estimated. Wetlands were therefore estimated to be a weak source of emissions which declined from 429 Gg Co2eq in 2000 to 267 Gg Co2eq in 2012. Wetlands included wetlands and water bodies.

the Settlements category (3B5) comprises the settlements remaining settlements and land converted to settlements subcategories. this category includes settlements and mines. Settlements remaining settlements do not have any Co2 emissions or uptakes, but land conversions to settlements produced a small source (1 195 Gg Co2 per year).

other lands (3B6) comprised emissions and removals from the biomass and soil carbon pools for other lands remaining other lands and land converted to other lands subcategories. the other lands category includes low shrublands, bare ground and degraded land. Other lands remaining other lands do not have any Co2 emissions or uptakes, while land converted to other lands was shown to be a sink of 960 Gg Co2 per year.

5.4.3 General methodology

South africa followed a tier 1 approach and in Forest lands a tier 2 approach of the ipCC 2006 Guidelines to determine the effects of land use and land-use change on the GHG inventory. the gain-loss method was used, where the inventory data was subdivided into the appropriate pools and land-use types.

it was assumed that no conversions occurred before 1990, so land converted to another land category would remain, and accumulate, in the converted category for 23 years. the ipCC default period is 20 years, but it is stated that if the change period extends beyond this then the time period can be adjusted. in this inventory the maps spanned a 23 year period therefore D = t = 23 years.

the inventory comprises the six land classes recommended by the ipCC (forest land, grassland, cropland, settlements, wetlands and other) as well as all the land-conversion categories. Due to the diverse nature of South africa’s vegetation, the six land classes had several subdivisions (Table 5.8). these divisions are further explained in Section 5.4.5.

the sub-divisions in this inventory differ slightly from those in the 2010 inventory. the main reason for this change was because in the previous inventory the maps suggested large changes between bare ground and Nama-Karoo and Succulent-Karoo vegetation. the fact that the biome borders used to identify these areas are static may havehadaninfluenceonthesenumbersandledtoerrorsin change over time. in this current inventory, the map resolution was increased from 1 km to 30m and the biome categorieswereremoved.Moredetailontheclassificationis provided in Section 5.4.5. area per category was estimated using a wall-to-wall approach, where the entire landsurfaceofSouthAfricawasclassifiedintooneofthesix land class. land areas were determined from the maps described in Section 5.4.4.

annual carbon stock changes in biomass were estimated using the process-based (gain-loss) approach where gains



are attributed to growth and losses are due to decay, harvesting, burning, disease, etc. For the land remaining in the same land-use category annual increases in biomass carbon stocks were estimated using equation 2.9 of the ipCC 2006 Guidelines, where the mean annual biomass growth was estimated using the tier 1 approach of equation 2.10 in the ipCC 2006 Guidelines. For plantations the tier 2 approach was applied. the annual decrease in carbon stocks due to biomass losses were estimated from equations 2.11 to 2.14 and 2.27 of the ipCC 2006 Guidelines. For the land converted to another land category, the tier 2 approach was used. Changes in biomass carbon stocks were estimated from ipCC 2006 equations 2.15 and 2.16. Carbon stock changes in dead organic matter

were estimated with the stock-difference method (tier 1), according to equation 2.19 and 2.23 of the ipCC 2006 Guidelines. Changes in mineral soil carbon stocks for both land remaining land and land converted to a new land use were estimated from the formulation B of equation 2.25 (ipCC, 2006 Guidelines, volume 4, p. 2.34).

if an ipCC default factor is required, the climate zone often plays a part in the selection process. For the purpose of this inventory all land classes were assumed to be within the warm temperate dry climate category as this is the dominant ipCC category for South africa based on long term climate data (Moeletsi et al., 2013).

table 5.8: Sector 3 aFolU – land: the six ipCC land classes and the South african sub-categories within each land class.

IPCC Land Class SA Sub-category in 2010 inventory Reclassification in this inventory

Forest land

Natural Forests indigenous Forests

plantations (eucalyptus, Softwoods, acacia, other sp.)

plantations/woodlots

thicket thicket/ Dense bush

Woodland/savannas Woodland/ open bush

Cropland

annual commercial cropsCultivated commercial annual crops (pivot, non-pivot)

perennial crops (viticulture, orchards)Cultivated commercial perennial crops (viti-culture, orchards)

annual semi-commercial/subsistence crops Cultivated subsistence crops

Sugarcane

GrasslandGrasslandFynbos

Grasslands

SettlementsSettlements Settlements

Mines Mines

WetlandsWetlands Wetlands

Water bodies Water bodies

other lands

Nama KarooSucculent KarooBare ground

low shrublandsBare groundDegraded land

other


5.4.4 Method for obtaining land-cover matrix

the South african National land-Cover Dataset 19903 and 2013-144, developed by Geoterraimage (Gti), were used for this study to determine long-term changes in land cover and their associated impacts. the 1990 and 2013-14 National land-Cover Datasets were derived from multi-seasonal landsat 5 and landsat 8 imagery with 30 x 30 m raster cells, respectively. the 1990 National land-Cover Dataset made use of imagery from 1989 to 1991, while the 2013-14 National land-Cover Dataset used 2013 to 2014 imagery. Both datasets were developed as 35 and 17 classes. the 35-class land cover, which included biome-level information (Fynbos, Nama-Karoo and Succulent-Karoo, where applicable) for indigenous forest, thicket/dense bush, Woodland/open bush, shrubland, grasslands and bare ground was further summarised into 17 classes to remove the biomes (see Table 5.7). Details of the land-cover datasets can be found in (Appendix G).

the accuracy of the 2013-14 land-Cover Dataset was calculated at 83% based on 6 415 sample points. it was determined that the accuracy is unlikely to be the result of chance occurrence, with a high Kappa score of 80.87. the 1990 dataset did not have an accuracy assessment conducted on it as there was no historical reference to use. the assumption was that the same modelling procedures were used to compile the 1990 dataset as was used for the 2013-14 dataset, therefore, the accuracy assessment calculated for the 2013-14 dataset would apply to the 1990 dataset.

landsat 5 and 8 imagery with a 30m resolution was acquired for the 1990 and 2013-14 datasets from the United States Geological Survey (USGS, http://glovis.usgs.gov/). Seasonal images were acquired to characterise seasonal variations of foundation-based landscapes, which include; trees, grass, water and bare ground. Spectral indices were derived from existing algorithms including,

the Normalised Difference vegetation index (NDvi), Normalised Difference Water index (NDWi) and Gti custom-derived algorithms. erDaS imagine © was used for all modelling. all modelling was conducted using the foundationclasses.Terrainmodificationswereconductedto minimise terrain-shadowing effects resulting from seasonalvariations.Thereafter,thespectrally-modifieddataset was merged into a single national dataset with the various classes. a detailed description of the modelling process can be obtained from the Gti report (2014, 2015) in Appendix G.


Details of the source imagery used to develop the land cover maps for South africa for 1990 and 2013-2014 are given in the Gti report (2014, 2015) provided in Appendix G.

5.4.4.2 Land Cover and Modelling Grid Resolution

the landsat 5 and 8 modelling was based on 30 x 30 m pixels,whilethegeographicmaskswerepre-definedinthe Gti data libraries.

5.4.4.3 Land Matrix Approach

the ipCC provides detailed guidelines for the three approaches to spatial data use (ipCC 2006, vol. 4, Chapter 3.3.1):

• Approach1identifiesthetotalareaforeachcategoryof land use within a country, but does not provide detailed information on conversions between different categories of land use;

• approach 2 incorporates approach 1, but introduces the tracking of conversions between land use categories; and

3 © GeoterraiMaGe - 2015

4 © GeoterraiMaGe - 2014



• approach 3 extends the information available from approach 2 by allowing land use conversions to be tracked in a spatially explicit way.

the maps developed for this inventory allowed for a tier 2 methodology (ipCC, 2006) to represent areas of land cover and change, and to develop the land cover-change matrix. the matrix was developed by intersecting the 1990 land-cover map with the 2013-14 land-cover map. all methodological details are provided in the Gti land-cover change report (2015) in Appendix G.

5.4.4.3.1 Land use changes

the resulting land use change matrix, including transitions, between 1990 and 2013-14 are shown in Table 5.9.

5.4.4.3.2Derivationofannualland-usechanges

the annual land cover changes between 1990 and 2013-2014 were calculated on a proportional basis, via linear interpolation, and are shown in Table 5.10. in the future additional data can be used to redistribute the annual change over the years.

table 5.9: Sector 3 aFolU – land: land-use matrix for the change between 1990 and 2013-14 (ha) with highlighted numbers on the diagonal showing the area (ha) remaining in the same category and the other cells show the relevant land-use changes (including the transitions since 1990).

Initial/ Final

Forest land

Crop-lands

Grass-lands

Settle-ments

Wet-lands

Other lands

Σ reduc-tions

Σ addi-tions – Σ reduc-tions

Forest land

14 275 143 443 513 2 824 017 185 925 105 633 2 119 071 5 678 158 3 075 446

Crop-lands

651 525 12 020 271 904 354 78 603 42 821 493 488 2 170 791 -221 986

Grass-lands

5 049 249 946 522 16 594 524 244 892 184 331 4 471 448 10 896 442 -1 696 993

Settle-ments

146 716 51 986 149 644 2 641 451 6 621 38 259 393 226 202 576

Wet-lands

291 891 79 294 442 552 6 994 2 642 354 265 096 1 085 826 -656 661

Other lands

2 614 223 427 490 4 878 884 79 388 89 759 48 441 893 8 089 743 -702 381

Σ addi-tions

8 753 603 1 948 804 9 199 449 595 801 429 165 7 387 362

Σ land use cate-

gory23 028 746 13 969 075 25 793 973 3 237 252 3 071 518 55 829 255

Total area

124 929 820


5.4.4.4 Limitations and Cautionary Notes on Change Assessment

Cloud-free images from landsat were used as far as possible. Where this was not possible, regions were corrected by merging cloud affected yoa with cloud free toa. the process of masking was found to be acceptable by Gti.

the seasonal representation and range of monthly acquisition dates is generally better for the 1990 land-cover dataset, than used for the 2013-14 dataset, despite the same average number of acquisition dates per frame for the two datasets. this is because the 2013-14 dataset was limited to what cloud free imagery was recorded between april 2013 and the designated March 2014 cut-off, whereas the 1990 dataset had access to a wider range of imagery from 1989 to 1993. this greater seasonal representationmayhaveinfluencedthemodellingresultsin terms of the distribution and extent of some of the natural vegetation classes (excluding indigenous forests), sinceabetterseasonalprofilewasoftenpossiblein1990compared to 2014 in some image frame locations.

the 1990 period appears to have been generally wetter than the 2013-14 period in most regions, as observed through the increase in observable surface water features in 1990. this is especially evident in the pans across the central regions of the Free State and Northern Cape, wheremanypansarewaterfilledin1990anddryin2013-

14.Thesedifferenceswillreflectaschangesbetweenthetwo assessment periods but do not necessarily represent a permanent loss of water bodies over the ± 24 year period, but rather a seasonal (or climatic) induced difference. this fact needs to be noted during modelling and interpretation of land-cover change results.

the wetter conditions prevalent in the 1990 dataset have allowed the wetland modelling to be extended across the full national coverage, unlike the drier 2013-14 image data which restricted wetland modelling to the wetter eastern andsouthernregions.Thesedifferenceswillreflectaschanges between the two assessment periods but do not necessarily represent a permanent loss of wetlands over the ± 24 year period, but rather a seasonal (or climatic) induced difference. this fact needs to be noted during modelling and interpretation of land-cover change results.

the wetter conditions prevalent in the 1990 dataset may haveinfluencedtheboundarybetweenlow-shrub(i.e.Karoo type) and Highveld grassland vegetation, since the 1990 grassland/low shrub transition in the western Free State region appears to have shifted generally to the west compared to 2013-14, which may be a result of a greater grass component being present in the Karoo/grass transition zone in 1990.

the (Highveld, montane and coastal) grassland correction procedures applied to the 1990 dataset are considered to be slightly more accurate than those applied to the

table 5.10: Sector 3 aFolU – land: annual land cover changes between the various land classes.

Initial/ Final Forest land Croplands Grasslands Settlements Wetlands Other lands

Forest land 148 115 122 783 281 627 4 593 22 005

Croplands 28 327 39 320 110 965 1 862 4 735

Grasslands 219 533 41 153 10 647 8 014 194 411

Settlements 6 379 7 811 6 506 288 741

Wetlands 12 691 13 897 19 241 34 440 7 421

Other lands 113 662 121 434 212 125 341 739 3 903



2013-14data,asaresultofbothmodifiedproceduresusedand better image seasonal representation in 1990. it is therefore potentially possible that some of the ‘grassland’ to ‘open woody vegetation’ changes in these regions may be a result of the improved grassland delineation in 1990 and not actual bush cover increases in 2013-14.

Changes in the extent of urban areas between 1990 and2013-14maynotbeaslocallysignificantasmaybeexpected, since the totalbuilt-upfootprintdefinedinthe“settlements” class in both the 1990 and 2013-14 datasets includes peripheral smallholding areas around the main built-upareas;andthesetendtobethefirstland-usethat is converted to formal urban areas, before further expansion into natural and cultivated lands.

it should also be noted that all land conversions as provided by the land change map were included due to the limited time period to assess the change maps. in the next inventory a full assessment of the land change classes will be completed.

5.4.4.5 Comparison with Land Cover Data for 2010 Inventory

For the purpose of this inventory a comparison was made between the 2010 national land cover data from the previous inventory (Department of environmental affairs, 2014) and the new 2013-14 land cover map used in this inventory (table 5.11). Croplands, settlements and other landareasweremuchlargerinthe2010classification,while forest land area was reduced by 44.3%. indigenous forest area remained fairly consistent, while the thicket/densebushandwoodland/openbushvariedsignificantly.Thisispartlybecauseofthere-classificationoftheclassesand the removal of the biome boundaries. plantation areas were reduced, which was expected as plantations were overestimated in the 2010 map. the map area for plantations is now more consistent with areas reported by Forestry South africa (FSa, 2012). Grassland areas are consistent, but the removal of the Fynbos boundary resultedintheFynbosbeingreclassifiedintotheother

land classes and it was thus moved out of the grassland category.AportionoftheFynboswasclassifiedintolowshrubland, which is partly the reason for the increase in the Other land category.

5.4.4.6 QA/QCandVerification

Gti calculated the accuracy of the 2013-14 land-cover dataset with the multi-seasonal landsat 8 imagery. the results indicated an overall map and mean map accuracy of 82.53% and 88.36%, respectively. a total of 6 415 sample points were used to determine the accuracy. in addition, the Kappa index of 80.87 indicates that results are unlikely to be from a chance occurrence. all classes recorded above 70% user accuracy (Ua) and producer accuracy (pa) except for dense bush/thicket (53.74% Ua), woodland/open bush (60.84% Ua; 54.13 pa), low shrubland/other (61.82 pa) and grasslands (69.82 pa). Detailed results can be found in the Gti report (Appendix G).

5.4.4.7 Time-series Consistency

it is important that there is a consistent time-series in the preparation of the land-use category and conversion data so that any artefact from method change is not included as an actual land-use conversion. When using land-use dataitisimportanttoharmonizedefinitionsbetweenthe existing independent data sources; ensure that data acquisition methods are reliable and well documented; ensuretheconsistentapplicationofcategorydefinitionsbetween time periods; prepare uncertainty estimates; and ensure that the national land area is consistent across the inventory time series. this inventory has tried to address these issues.

landsat imagery was used for both the 1990 and 2013-14 datasetstospecificallyensuretime-seriesconsistency.Inaddition, the resolution of landsat imagery was improved compared to the MoDiS datasets used in previous studies. the methods and procedures used to develop the 1990 and 2013-14 datasets remained consistent. the


tabl

e 5.

11:

Sect

or 3

aFo

lU –

lan

d: C

ompa

riso

n be

twee

n th

e la

nd a

reas

from

the

201

0 in

vent

ory

and

the

upda

ted

area

s in

thi

s 20

12 in

vent

ory.

IPC

C

cate

gory

Vege

tati

on c

lass

(20

10)

Are

a (h

a)Ve

geta

tion

cla

ss (

2013

-14)

Are

a (h

a)C

hang

e re

lati

ve t

o 20

10 (

%)

Fore

st

Land

indi

geno

us F

ores

t47

7 88

9in

dige

nous

For

est

428

444

t

hick

et2

876

607

thi

cket

/Den

se b

ush

8 29

1 66

9

Woo

dlan

ds/S

avan

na35

854

791

Woo

dlan

d/o

pen

bush

12 4

34 9

32

plan

tatio

ns2

146

347

plan

tatio

ns /

Woo

dlot

s 1

873

701

Su

b-to

tal

41 3

55 6

3423

028

746

-44.

32%

Cro

plan

d

ann

ual c

omm

erci

al c

rops

non

-piv

ot6

445

523

Com

mer

cial

ann

ual c

rops

non

-piv

ot10

610

838

ann

ual c

omm

erci

al c

rops

piv

ot42

6 97

3C

omm

erci

al a

nnua

l cro

ps p

ivot

782

049

perm

anen

t cr

ops

orch

ard

181

835

Com

mer

cial

per

man

ent

orch

ards

346

950

perm

anen

t cr

ops

vitic

ultu

re23

5 11

3C

omm

erci

al p

erm

anen

t vi

nes

188

711

ann

ual s

emi-c

omm

erci

al /

subs

iste

nce

crop

s1

020

776

Cul

tivat

ed s

ubsi

sten

ce c

rops

2 04

0 52

7

Suga

rcan

e31

3 61

3

Sub-

tota

l8

623

833

13 9

69 0

7561

.98%

Gra

ssla

ndG

rass

land

24 8

52 6

54G

rass

land

s25

793

973

Fynb

os5

909

183

Sub-

tota

l30

761

837

25

793

973

-16.

15%

Sett

le-

men

tsSe

ttle

men

ts1

822

753

Sett

lem

ents

(in

cl. s

mal

lhol

ding

s)2

908

280

M

ines

17

9 32

4M

ines

328

973

Su

b-to

tal

2 00

2 07

7

3 23

7 25

261

.69%

Wet

land

sW

etla

nds

1 13

8 32

2W

etla

nds

1 02

5 90

0

Wat

er B

odie

s1

716

648

Wat

er b

odie

s2

045

618

Su

b-to

tal

2 85

4 97

0

3 07

1 51

87.

58%

Oth

er

land

s

Bare

Gro

und

22 6

63 3

54Ba

re G

roun

d13

057

933

o

ther

922

077

Nam

a-K

aroo

9 28

0 57

3lo

w s

hrub

land

41 8

27 2

60

Succ

ulen

t K

aroo

3 59

8 41

0

D

egra

ded

944

061

Su

b-to

tal

36

464

414

55

829

255

53.1

1%



main variation in the development of the two land-cover datasets was the use of landsat 5 and landsat 8. all other aspects in this development remained consistent.

5.4.4.8 Planned Improvements

the land-cover maps were developed as a desktop only modelling exercise, the results of which are directly dependent on the validity and accuracy of the modelling data inputs, theoretical assumptions and associated modellingrules.Assuch,nostatisticalverificationoffinalland-cover change detection accuracy can or has been be provided. the Dea has a GHG inventory improvement Programmeand,althoughtherearenospecificplansin place currently, it is recommended that under this programmeastratifiedsamplingapproachwhichcanbeused to validate these, and future, land-cover maps be investigated for future inventory updates.

5.4.5 Land use definitions and classifications

5.4.5.1 Forest Land

the forest land category included all land with woody vegetationconsistentwiththresholdsusedtodefineforest land in the national greenhouse gas inventory. it also included systems with a vegetation structure that currently falls below, but in situ could potentially reach thethresholdvaluesusedbyacountrytodefinetheforestland category. according to the Kyoto protocol and the MarrakeshAccord,aforestisdefinedashavingaminimumarea of 0.05 to 1.0 hectares with tree crown cover (or equivalent stocking level) of more than 10 to 30%, with trees with the potential to reach a minimum height of 2 to 5 metres at maturity in situ. a forest may consist either of closed forest formations where trees of various storeys and undergrowth cover a high proportion of the ground or open forest. young natural stands and all plantations which have yet to reach a crown density of 10 to 30% or tree height of 2 to 5 metres are included under forest.

TheFAOdefinesindigenousforestsasforestswithatreecanopy cover of more than 10%, trees should be able to reach a height of 5 m and the forest area must be greater than 0.5 hectares. Under this definition, indigenous forests include woodlands which have a tree canopy cover of 10 to 70%. in the Global Forest resources assessment report for South africa (GFraSa, 2005) forests include (a) forests with a tree canopy cover >70%, (b) forest and woodlands which have a tree cover of between 10% to 70%,and(c)plantations.Thisreportthereforeclassifiesallwoodlands and savannas under forest land. the previous inventory differed in that it separated out the woody savannas from other woodlands by their geographical distribution, and excluded woody savannas from forest lands.

a further addition to forest land in this inventory was thickets. typically thickets are not classed as forest lands as the trees do not reach the height requirement. in the GFRASA(2005)theywereclassifiedasotherwoodlands,however, for the purpose of this inventory, thickets were classed as forest land, since they are more like forests than grasslands.

the following vegetation subcategories were therefore included within forest land:

• Natural forests:

o Natural/semi-natural indigenous forest , dominated by tall trees, where tree canopy heights are typically > ± 5m and tree canopy densities are typically > ± 75 %, often with multiple understory vegetation canopies.

• plantations:

o all areas of systematically planted man-managed tree resources, composed of primarily exotic species, including hybrids. this category includes all plantations from young to mature which have been established for commercial timber production seedling trials or woodlots (thompson, 1999; Geoterraimage, 2013);


o it includes clear-felled stands;

o it excludes all non-timber-based plantations, as well as orchards.

• thickets/dense bush:

o Natural/semi-natural tree- and/or bush-dominated areas, where typically canopy heights are between 2 –and 5 m, and canopy density is typically > ± 75%, but may include localised sparser areas down to ± 60%. includes dense bush, thicket, closed woodland, tall, dense shrubs, scrub forest and mangrove swamps. Can include self-seeded bush encroachment areas if sufficientcanopydensity.

• Woodland/open bush:

o Natural/semi-natural tree- and/or bush-dominated areas, where typically canopy heights are between ± 2 and 5 m, and canopy densities typically between 40 and 75%, but may include localised sparser areas down to ± 15 to 20%. includes sparse, open bushland and woodland, including transitional wooded grassland areas. Can include self-seeded bush encroachment areas if canopy density is within indicated range.

5.4.5.2 Cropland

this category included cropped land and agro-forestry systems where the vegetation structure falls below the thresholds used for the forest land category. the subcategories in croplands were:

• annual commercial crops:

o Commercial annuals (rain fed or non-pivot) - Cultivated lands used primarily for the production of rain-fed, annual crops for commercial markets. typically represented by largefieldunits,oftenindenselocalorregionalclusters.Inmostcasesthedefinedcultivatedextent represents the actual cultivated or potentially extent. includes Sugarcane crops.

o Commercial pivot (irrigated or pivot) - Cultivated lands used primarily for the production of centre pivot irrigated, annual crops for commercial markets.Inmostcasesthedefinedcultivatedextent represents the actual cultivated or potentially extent. includes sugarcane crops.

• perennial crops:

o permanent orchards - Cultivated lands used primarily for the production of both rain-fed and irrigated permanent orchard crops for commercial markets. includes both tree, shrub and non-woody crops, such as citrus, tea, coffee, grapes, lavender and pineapples etc. in most casesthedefinedcultivatedextentrepresentsthe actual cultivated or potentially extent.

o permanent vines - Cultivated lands used primarily for the production of both rain-fed and irrigated permanent vine (grape) crops for commercial markets. in most cases the definedcultivatedextentrepresentstheactualcultivated or potential extent.

• Subsistence crops:

o those that do not meet the threshold for annual commercial crops and are found in and around local housing areas;

o Cultivated lands used primarily for the production of rain-fed, annual crops for local markets and / orhomeuse.Typicallyrepresentedbysmallfieldunits, often in dense local or regional clusters. Thedefinedareamayincludeintra-fieldareasof non-cultivated land, which may be degraded oruse-impacted,iftheindividualfieldunitsaretoosmalltobedefinedasseparatefeatures.

5.4.5.3 Grassland

the grassland category included range and pasture lands that were not considered cropland. the category also included all grassland from wild lands to recreational areas as well as agricultural and silvo-pastural systems,



consistentwithnationaldefinitions.Itisdefinedasnatural/semi-natural grass-dominated areas, where typically the tree and/or bush canopy densities are < ± 20 %, but may include localised denser areas up to ± 40%, (regardless of canopy heights). includes open grassland, and sparse bushland and woodland areas, including transitional wooded grasslands. May include planted pasture (i.e. grazing)ifnotirrigated.“Plantedgrassland”isdefinedin the same way except that it is grown under human-managed (including irrigated) conditions for grazing hay or turf production or recreation. planted grassland can be either indigenous or exotic species. irrigated pastures willtypicallybeclassifiedascultivated,andurbanparksand golf courses etc. under urban.

5.4.5.4 Settlements

Settlementsaredefinedasalldevelopedland,includingtransportation infrastructure and human settlements of any size, unless they are already included under other categories. they essentially comprise all formal built-up areas in which people reside on a permanent or near-permanentbasisidentifiablebythehighdensityofresidential and associated infrastructure. includes both towns villages and where applicable the central nucleus of more open rural clusters. this category also includes mines. the mining activity footprint includes extraction pits,tailings,wastedumps,floodedpitsandassociatedsurface infrastructure such as roads and buildings (unless otherwise indicated), for both active and abandoned mining activities. this class may also include open cast pits, sand mines, quarries and borrow pits etc.

5.4.5.5 Wetlands

Wetlands are areas of land that are covered or saturated by water for all or part of the year and that do not fall into the forest land, cropland, grassland or settlements categories. they include reservoirs as a managed sub-division and natural rivers and lakes as unmanaged sub-divisions. Wetlands include two sub-divisions:

• Water bodies:

o areas occupied by (generally permanent) open water. the category includes both natural and man-made water bodies that are either static orflowingandfreshbrackishandsalt-waterconditions (thompson, 1999). this category includes features such as rivers, major reservoirs, farm-level irrigation dams, permanent pans lakes and lagoons;

• Wetlands:

o Natural or artificial areas where the water level is permanently or temporarily at (or very near) the land surface typically covered in either herbaceous or woody vegetation cover. the vegetationcanbeeitherrootedorfloating.Wetlands may be either daily (i.e. coastal), temporarily seasonal or permanently wet and/or saturated. vegetation is predominately herbaceous. includes, but is not limited to, wetlands associated with seeps/springs, marshes, floodplains,lakes/pans,swamps,estuaries,andsome riparian areas.

5.4.5.6 Other Land

this category includes bare soil, rock, ice, and all land areasthatdidnotfallintoanyoftheotherfivecategories.Itallowedthetotalofidentifiedlandareastomatchthenational area, where data were available. this category included the following:

• Bare ground:

o Bare, non-vegetated ground, with little or very sparse vegetation cover (i.e. typically < ± 5 to 10 % vegetation cover), occurring as a result of either natural or man-induced processes. includes, but is not limited to, natural rock exposures, dry river beds, dry pans, coastal dunes and beaches, sand and rocky desert areas, very sparse low shrublands and grasslands, erosion areas, and major road networks etc.


Mayalso include long-termwildfirescars insome mountainous areas in the Western Cape.

• Degraded land:

o Sparsely vegetated areas, occurring as a result of man-inducedprocesses,whichshowsignificantlylower overall vegetation cover compared to surrounding, natural undisturbed areas. includes, but is not limited to, natural rock exposures, dry river beds, dry pans, coastal dunes and beaches, sand and rocky desert areas, very sparse low shrublands and grasslands, erosion areas, and major road networks etc.

• low shrublands:

o Natural/semi-natural low shrub-dominated areas,typicallywith≤2mcanopyheight.Includesa range of canopy densities encompassing sparse to dense canopy covers. very sparse covers may be associated with the bare ground class. typically associated with low, woody shrub, Karoo-type vegetation communities, although can also represent locally degraded vegetation areas where there is a significantly reduced vegetation cover in comparison to surrounding, less impacted vegetation cover, including long-termwildfirescarsinsomemountainousareasin the Western Cape. Note that taller tree/bush/shrub communities within this vegetation type aretypicallyclassifiedseparatelyasoneoftheother tree- or bush-dominated cover classes.

5.4.5.7 Managed and Unmanaged Lands

AccordingtotheIPCCdefinitionoftheAFOLUsector,anthropogenic GHG emissions and removals take place on “managed land”. Most land in South africa is managed to some degree, so all land classes were included in the biomass growth and loss calculations, except other lands. inclusion of the other land category may be considered in the future.

5.4.6 Recalculations

there were several changes and thus recalculations for the land-use sector since the last inventory:

• Newland-covermaps:

o these were developed specifically for this inventory to cover the period 1990 – 2013-14;

o the resolution of the maps was increased from 1km to 30m.

• Reclassificationoflandclasses:

o in this inventory the biome boundaries were removed so that Karoo and Fynbos vegetation were not distinguished. instead all vegetation fromtheseclasseswereclassifiedintotheothervegetation categories.

o in the previous inventory Fynbos vegetation was classed in the grassland category, but with the newreclassificationanddistributionofFynbosinto the other land categories, the grassland category consisted only of grasslands.

o Sugar cane was incorporated into annual crops.

o a small degraded land category was included.

• land-conversion data:

o the category other lands was also included and sinks and removals from this category were incorporated in this inventory.

o Converted land was accumulated over the 23 year period.

• activity data:

o Updated biomass data for indigenous forests, woodland/open bush, orchards and vineyards were incorporated;

o Cropland methodology was corrected and updated with new biomass data;

o DoM was included for forest lands; and



o Soil type was included into the soil carbon data.

• HWp:

o emissions and removals from the HWp category were corrected and the methodology was updated following the recent Kp Supplementary GpG.

Specificrecalculationscompletedforeachlandclassarediscussedindetailunderthespecificsectionsbelow.Allrecalculations were completed for the years 2000 to 2012.

5.4.7 Forest land [3b1]

5.4.7.1 Source-Category Description

reporting in this category covers emissions and removals from above-ground and below-ground biomass, DoM and mineral soils. the category included indigenous forests, plantations/woodlots, thickets/dense bush, and woodlands/open bush. as in the previous inventory the plantations were sub-divided into eucalyptus sp., softwood sp., acacia (wattle) and other plantation species. Softwoods were further divided into sawlogs, pulp and other as the growth and expansion factors of these plantations differed. the majority of the Eucalyptus plantations are used for pulp so the eucalyptus species were not split by use, but Eucalyptus grandis and Other Eucalyptus species were separated.

Changes in biomass include wood removal, fuelwood collection, and losses due to disturbance. Harvested wood was included for plantations, while fuelwood collection was estimated for all forest land subcategories. in plantations,disturbancefromfiresandotherdisturbanceswas included, while for all other subcategories only disturbancefromfirewasincludedduetoalackofdataon other disturbances. it should be noted that only Co2 emissionsfromfireswereincludedinthissectionasallother non-Co2 emissions were included under section 3C1. also all emissions from the burning of fuelwood for

energy or heating purposes were reported as part of the energy sector.

this category reports emissions and removals from the categories forest land remaining forest land and land converted to forest land (new forest established, via afforestation or natural succession, on areas previously used for other land-use classes). in 2012 the total sink from forests amounted to 34 325 Gg Co2, which was up from the 21 565 Gg Co2 in 2000. land converted to forest land contributed the largest portion to the forest category (Figure 5.13) due to changes in biomass and soil carbon.Theannualfluctuationsinthiscategoryareduemainly to changes in biomass losses (due to harvesting, fuelwood removals and disturbance losses). in 2004 there was a reduction in the sink compared to 2003 and this was due mainly to an increase in disturbance losses that year. Harvested wood was also higher in 2004 compared to 2003. there was also an increase in losses due to disturbance in 2007 and 2008, which declined in 2009. these increases in carbon losses led to a sink in the forest land remaining forest land during 2008. this sink is however also partly due to the inability to allocate carbon losses between land remaining land and converted land. the increased sink in 2012 is due to a combination of reduced harvested wood, fuelwood and disturbance. Conversion of grasslands and other lands to forest lands are estimated to be the largest contributors to the land converted to forest land Co2 sink. of the 2012 sink, 90.2% was contributed by the biomass pool, while the rest was from the SoC and DoM pools (Figure 5.14).


Figure 5.14: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the forest land between 2000 and 2012, differentiated by source category.

Figure 5.13: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in the forest land between 2000 and 2012, differentiated by sub-category.




5.4.7.2.1 Biomass

Forest land remaining forest land

removals and emissions of Co2 from changes in above- and below-ground biomass are estimated using the tier 1 gain-loss Method in the 2006 ipCC Guidelines. the gains in biomass stock growth were calculated using the following equations (equation 2.9 and 2.10 from ipCC 2006 Guidelines):

ΔCG=∑(Ai * Gtotali * CFi)

(eq. 5.14)

Where for Gtotal a tier 1 approach was used for natural vegetation classes (eq. 5.15), while a tier 2/3 approach was applied to the plantations (eq. 5.16)

Gtotali=∑[GWi * (1+r)]

(eq. 5.15)

Gtotali=∑[Iv * BCeFi * (1+r)]

(eq. 5.16)

and:

ai = area of forest category i remaining in the same land-use category

GWi = average annual above-ground biomass growth for forest category i (t dm ha-1 a-1)

ri = ratio of below-ground biomass to above- ground biomass for forest category i (t dm below-ground biomass (t dm above-ground biomass)-1)

Iv =averagenetannualincrementforspecific vegetatio type (m3 ha-1 yr-1);

BCeFi = biomass conversion and expansion factor for conversion of net annual increment in volume to above-ground biomass growth ((t aG biomass growth (m3 net annual increment)-1);

CFi = Carbon fraction of dry matter for forest category i (t C (t dm)-1)

in the previous inventory a dead wood component was included in eq. 5.15 as it was indicated that the carbon losses from the live biomass would be overestimated as the fuel wood removals include dead wood. in this inventory estimates for the DoM pool were included therefore, instead of including a dead wood component in the carbon gains the fuelwood removals were adjusted for dead wood so only removals of live wood were included here.

the losses were calculated for three components:

• loss of carbon from harvested wood;

• loss of carbon from fuelwood removals; and

• loss of carbon from disturbance.

loss of carbon from harvested wood was calculated for plantations only as follows (equation 2.12 ipCC 2006 Guidelines):

lwood-removals = [H * BCeFS * (1+r) * CF

(eq. 5.17)

Where:

H = annual wood removals (m3 yr-1)

BCeFS = biomass conversion and expansion factor for conversion of removals in merchant- able volume to total biomass removals (including bark), (t biomass removed (m3 of removals)-1)


r = ratio of below-ground biomass to above- ground biomass (t dm below-ground (t dm above-ground)-1).

CF = Carbon fraction of dry matter (t C (t dm)-1)

loss of carbon from fuelwood removals was calculated using the following equation (equation 2.13 of ipCC 2006 Guidelines):

lfuelwood = [FGtrees *BCeFr * (1+r) + FGpart * D] * CF

(eq. 5.18)

Where:

FGtrees = annual volume of fuelwood removal of whole trees (m3 yr-1)

FGpart = annual volume of fuelwood removal as tree parts (m3 yr-1)

BCeFr = biomass conversion and expansion factor for conversion of removals in merchant- able volume to biomass removals (including bark), (t biomass removal (m3 of removals)-1)

r = ratio of below-ground biomass to above- ground biomass (t dm below-ground (t dm above-ground)-1)

D = basic wood density (t dm m-3)

CF = carbon fraction of dry matter (t C (t dm)-1)

Finally, the loss of carbon from disturbance in plantations was calculated following ipCC equation 2.14:

ldisturbances = adisturbance * BW * (1+r) * CF * fd

(eq. 5.19)

Where:

adisturbance = area affected by disturbances (ha yr-1)

BW = average above-ground biomass of areas affected by disturbance (t dm ha-1)

r = ratio of below-ground biomass to above- ground biomass (t dm below-ground (t dm above-ground)-1).

CF = carbon fraction of dry matter (t C (t dm)-1)

fd = fraction of biomass lost in disturbance; a stand-replacing disturbance will kill all (fd = 1) biomass while an insect disturbance may only remove a portion (e.g. fd = 0.3) of the average biomass C density

For woodlands/ open bush the loss of carbon from disturbance(firesonly)wascalculatedusingthebiomassburning equation (ipCC 2006, equation 2.27, p. 2.42):

lfire = a * MB * Cf * Gef *10-3 * (12/44)

(eq. 5.20)

Where:

lfire =massofcarbonemissionsfromfire(tC);

a = area burnt (ha)

MB = mass of fuel available for combustion (t ha-1)

Cf = combustion factor

Gef = emission factor (g Co2 (kg dm burnt)-1)

this ensured that the disturbance estimates were consistent with the biomass burning estimates.



Thetotalcarbonflux(ΔC)wasthencalculatedasfollows(ipCC 2006 equations 2.7 and 2.11):

ΔC=ΔCG–Lwood-removals – lfuelwood – ldisturbances

(eq. 5. 21)

Land converted to forest land

the land converted to plantations could not be split into the various plantation types due to a lack of data so the values for eucalyptus were used as this is the dominant species. annual increase in biomass carbon stocks were calculated in the same way as for forest land remaining forest land. if a land area is converted to a plantation then it is unlikely wood would be harvested in the year following the conversion, therefore it was assumed that no wood was harvested or removed for fuel wood fromtheconvertedlandinthefirstyearofconversion.all annual harvested wood and fuel wood losses were therefore allocated to the forest land remaining forest land category.Therewasinsufficientdatatodeterminehowmuch of the annual disturbance in a plantation occurred on the converted land. For the purpose of this inventory it was assumed that no disturbance occurred on converted land, but this assumption will need to be corrected in futureinventories.Lossesduetofiredisturbancewerecalculated for land converted to woodlands/open bush using equation 2.27 in the ipCC 2006 Guidelines. a detailed description of the methodology for calculating lossesduetofireisgiveninSection 5.5.2. as with forest land remaining forest land, indigenous forests and thickets were assumed not to burn.

5.4.7.2.2Deadorganicmatter

Forest land remaining forest land

the dead organic matter (DoM) pool comprises deadwood and litter. For this inventory a tier 1 approach was introduced as the DoM pool is not a key category.

Under the tier 1 method the DoM pool in land remaining in the same land use category are reported to have zero changes in carbon stocks or carbon emissions (ipCC, 2006, Chapter 2, p 2.21).


For the tier 1 approach it is assumed that on land converted to forest land the build-up of litter and deadwood occurs linearly from zero (in the non-forest land-use categories) over a transition period. the transition period for this inventory was taken as 23 years (transition period of the land-cover maps), as ipCC Guidelines indicate that if the transition time (t) is longer than the default 20 years (D) then the value for t should be used. annual change in carbon stocks in deadwood or litter (tonnes C yr-1) was calculated using the following equation (equation 2.23 in the ipCC 2006 Guidelines):

ΔCDoM = [(Cn – Co) * aon] / ton

(eq. 5.22)

Where:

Cn = dead wood/litter stock under the new land-use category (tonnes C ha-1);

Co = dead wood/litter stock under the old land-use category (tonnes C ha-1);

aon = area undergoing conversion from old to newland use category (ha);

ton = time period of the transition from old to new land-use category (yr.) (23 years for this inventory).


5.4.7.2.3 Mineral soils


annual change in carbon stocks in mineral soils were calculated by applying a tier 1 method with equation 2.25 of the ipCC 2006 Guidelines (ipCC, 2006, volume 4, p. 2.30).Landareaswerestratifiedbydefaultsoiltypes(seeSection 5.4.1) in order to obtain SoC reference values (for warm temperate dry conditions). these were then multiplied by stock change factors by using formulation B of equation 2.25:

ΔCMineral=∑[{(SOCreF*FlU*FMG*Fl)0 – (SoCreF*FlU*FMG*Fi*a)(0-t)}*a]/D

(eq. 5.23)

Where:

SoCreF = the reference carbon stock (t C ha-1) for each soil type;

FlU = stock change factor for land-use system for a particular land-use (dimensionless);

FMG = stock change factor for management regime (dimensionless);

Fi = stock change factor for input of organic matter (dimensionless);

time 0 = last year of inventory time period;

time (0-t) = beginning of the inventory time period;

a = land area (ha);

D = time dependence of stock change factor (transition period of 23 years was used for this inventory).


the land areas for forest land remaining forest land and land converted to forest land were obtained from the land use maps discussed in section 5.4.4.

5.4.7.3.1 Biomass

biomass gains

annual biomass growth data for plantations was calculated from mean annual increment (Mai) data provided by Forestry Sa. these values were multiplied by the BCeF’s from Dovey and Smith (2005) (0.522 for softwoods; 0.558 for Eucalyptus grandis; 0.741 for other Eucalyptus species; 0.909 for Acacia species; and an average of 0.68 for other species) to obtain the increment in t dm ha-1. these BCeFi

values fall within the ipCC 2006 default range of 0.4 – 0.9 t biomass (m3 wood volume)-1 for dry subtropical plantations with a growing stock between 100 and 200 m3 (ipCC 2006, table 4.5).

Midgley and Seydack (2006) reported that growth was 1% of aGB for indigenous forests. the aGB value in the previous inventory was 81 t dm ha-1 (ipCC GpG 2003), however in this inventory a value of 150 t dm ha-1 was applied based on the outputs of the recently completed NtCSa (Department of environmental affairs, 2015) (see Appendix F). the report provided validation for these data outputs. this value is at the upper limit of the ipCC default values for subtropical systems in africa (20 – 200 t dm ha-1) with the subtropical dry forest having a value of 130-140 t dm ha-1 (ipCC, 2003). the 1% growth rate of 1.5 t dm ha-1 yr-1 is also much more in line with ipCC default values for subtropical systems (table 4.12; ipCC 2006 Guidelines) than the 0.8 t dm ha-1 from the previous inventory.

Data for thickets remain unchanged from the previous inventory. Carbon sequestration rates of some thicket species have been estimated to be between 1.2 and 5.1 t C ha-1 yr-1 (2.4 to 10.2 t dm ha-1 yr-1) (aucamp and



Howe, 1979; Mills and Cowling, 2006; van der vyver, 2011). Mills and Cowling (2006) reported growth rates for above ground biomass of 0.04 to 0.131 g C m-2 yr-1 (or 0.85 to 2.76 t dm ha-1) in thickets which provides an average estimate of 1.8 t dm ha yr-1 as used in the previous inventory. For woodland/open bush a growth value of 0.9 t ha-1 yr-1 was also taken from the previous inventory (Department of environmental affairs, 2013).

a root-to-shoot ratio of 0.24 for woodland/open shrubs and forests taken from the Global Forest resource assessment for South africa (GFraSa, 2005), as was the 0.34 for acacia and other plantation species. For eucalyptus and softwoods the value of 0.15 was taken from Christie and Scholes (1995). all of these are consistent with the values used in the 2010 inventory (Department of environmental affairs, 2014). in the GFraSa (2010) thickets are classed as other woodlands and these are indicated to have a root-to-shoot ratio of 0.48.

the ipCC 2006 default value of 0.47 t C per t dm-1 (ipCC 2006, table 4.3) was used for the carbon fraction of dry matter of all Forest lands.

Losses due to wood harvesting

loss of carbon due to wood harvesting was only determined for plantations using FSa data (FSa, 2012) (Table 5.12) as wood harvesting does not occur in woodlands/open bush, thickets or indigenous forests. the wood harvested for fuelwood was not included here as it was included under fuelwood removals. the industry conversion factors provided were used to convert between tonnes and m3. the BCeFS were taken to be the same as the BCeFi values described above Section as the ipCC 2006 default values (ipCC 2006, table 4.5) indicate this to be the case.

table 5.12: Sector 3 aFolU – land: annual wood harvest data (m3/yr) from plantations between 2000 and 2012 (source: FSa, 2012).

Softwood-sawlogs

Softwood – pulp

Softwood – other sp.

Euc. grandis

Other Eucs

Wattle Other sp.

2000 4 086 816 3 574 970 147 184 4 510 395 2 951 744 1 205 401 34 992

2001 4 041 832 3 058 964 126 713 4 695 726 3 466 830 953 701 20 742

2002 4 041 765 3 126 909 135 942 5 035 023 2 993 119 992 205 60 023

2003 4 857 374 3 402 798 122 055 5 471 160 4 155 741 960 728 18 665

2004 5 040 134 2 765 001 155 901 7 529 361 3 512 057 993 457 48 305

2005 5 224 656 3 751 951 199 081 7 529 566 3 893 000 1 084 651 27 658

2006 5 507 744 3 636 937 184 414 8 722 087 3 263 045 1 176 582 42 648

2007 5 115 671 3 759 857 127 443 6 399 489 3 659 702 954 447 7 515

2008 4 894 655 3 837 417 109 575 6 822 931 3 386 144 796 124 18 746

2009 4 145 537 3 421 600 94 450 6 837 481 3 125 173 928 456 21 027

2010 3 744 396 3 343 702 100 615 5 372 218 3 733 368 900 317 36 038

2011 3 916 832 3 534 944 106 087 5 757 557 3 779 164 1 070 608 48 537

2012 4 411 143 3 308 541 99 550 5 411 542 3 634 810 980 161 42 180


Losses due to fuelwood removal

Fuelwood removal was estimated for all sub divisions within the Forest land class. For plantations FSa (2012) providedannualdataonwoodremovedforfirewood/charcoal (Table 5.13).

For natural forests, woodlands and shrublands data at a national scale is limited. Fuelwood consumption, therefore, was calculated by obtaining an average fuelwood consumption rate per household (Shackleton, 1998; Shackleton & Shackleton, 2004; Madubansi & Shackleton, 2007; Matsika et al., 2013) and combining this with the number of households that use fuelwood (Statistics Sa). a small percentage of the wood collected is deadwood, so this amount needs to be removed from fuelwood consumption, as only live wood removals are to be accounted for in this section. Shackleton (1998) calculated that annual deadwood production was 1.7% of total biomass. it was assumed that all pre-existing

deadwood had been collected the previous year, so only the deadwood produced in the current year is collected. therefore deadwood is assumed to be 1.7% of fuelwood consumption. the fuelwood consumption numbers are within the range of the value provided by the Fao. the fuelwood consumption estimates show a decline since 2000, but Damm and triebel (2008) suggested that fuelwood consumption was stabilizing and may decline in the future. there is very little information on how this amount is split between the various vegetation types, therefore, the whole amount was assumed to be taken from woodlands/open bush with no removal from forests and thickets. For woodland/savannas the BCeF of 2.1 t biomass removed (m3 of removals)-1 was taken from the previous inventory. this was determined from the BeF and wood density data in the GFraSa (2005). this value is high compared to the plantation data, but little data is available to determine the BCeF for woodlands. Data is being collected to improve this value in the next inventory. this higher value could be leading to an underestimation

table 5.13: Sector 3 aFolU – land: annual removal of fuelwood (m3/yr) from plantations between 2000 and 2012 (source: FSa, 2012).

Softwoods Euc. grandis Other Eucs Wattle Other sp.

2000 18 855 42 082 18 933 80 944 5 165

2001 21 523 95 890 21 093 94 991 6 268

2002 22 356 56 000 34 234 113 718 2 427

2003 20 148 53 439 36 821 101 730 2 822

2004 24 333 78 488 61 713 99 400 998

2005 39 772 51 053 37 221 134 037 2 289

2006 42 661 82 002 35 939 130 150 292

2007 26 537 93 077 51 850 132 358 1 674

2008 28 396 75 649 44 144 122 114 0

2009 38 669 105 093 43 624 123 065 1 727

2010 21 051 94 642 37 541 127 498 0

2011 25 448 139 666 18 019 116 852 0

2012 23 654 144 275 18 568 118 228 0



of the forest sink. During wood collection the trees are not destroyed, but are left to continue to grow. therefore roots were not removed during fuelwood harvesting (i.e. root: shoot ratio was set to zero).

Losses due to disturbance

the only disturbance losses that were estimated for all forest land classes were those from fire, but for plantations the loss due to other disturbances was also included. the FSa provides data on the area damaged duringfireandotherdisturbances.Fortheyear1979to 2000 the FSa also recorded whether the plantations in a disturbed area were slightly damaged, seriously damage (in terms of timber sold) or seriously damaged in that there was a total loss of vegetation. this data was used to determine the weighted average fraction of biomass lost in the disturbance (by assuming a slight damage had an fd = 0.3 and serious damage with total loss had a fd = 1) for hardwood (fd = 0.636) and softwood (fd = 0.489) species. these averages were kept constant over the 10-year period as the FSa stop reporting this data in 2000. the aGB (BW) data obtained from the NtCSa (Department of environmental affairs, 2015) only provided a value for plantations as a whole (97 t dm ha-1). to estimate biomass data for each plantation type the growing stock (by volume) data from GFraSa (2010) was combined with the BCeFS (Chhabra et al., 2002). the values ranged between 70 t dm ha-1 and 132 t dm ha-1 which are within the range estimated in the NtCSa (Department of environmental affairs, 2015).

Theamountofcarbonlostduetofiredisturbanceinwoodlands/open bush was determined as discussed in Section 5.4.7.2.1. Natural forests and thickets were assumed not to burn. the MB, Cf and emission factors are provided and discussed in Section 5.5.2.

5.4.7.3.2Deadorganicmatter

For the tier 1 approach the default litter values should be considered, however these values are very high relative to estimates obtained from the literature for South africa. the NtCSa (Department of environmental affairs, 2015) provided the values of 121 ± 49 g C m-2, 900 ± 50 g C m-2 and 254 ± 52 g C m-2 for woodlands, forests and thickets respectively (Shea et al., 1996; Weider and Wright, 1995; powell, 2009). these values were used in the inventory as opposed to the default values.


The long termclimate inSouthAfricawasclassifiedaccording to the ipCC climate classes and 99.7% of South africa falls within the “warm temperate dry” category (Moeletsi et al., 2013). therefore there was no stratificationbyclimateinthisinventory.SoilswerealsoclassifiedaccordingtothesixIPCCclasses(Moeletsiet al., 2013), so the land classes (for 1990 and 2013/14) werestratifiedbysoiltypesothatthereferencecarbonstock (SoCreF) could be determined. as in the previous inventory, the ipCC default SoCreF values were applied (ipCC 2006, volume 4, p. 2.31, table 2.3), however, the next inventory should include country-specific SoC reference values. the soil carbon values for the various vegetation types determined from the NtCSa (Department of environmental affairs, 2015),which ranged between 3.8 and 37.95 t C ha-1, were compared to the ipCC SoC reference values and found to be in a similar range. reference values are expected to be slightly higher as it is the carbon in natural systems where no conversions have taken place. For forest lands, grasslands, settlements and other lands the soil C stocks are assumed equal to the reference value (i.e. FlU, FMG and Fi = 1). Stock change factors were determined for each crop type by using data reported in Moeletsi et al. (2013) and the NtCSa (Department of environmental affairs, 2015) (Table 5.14). For settlements a stock change factor of 0.8 was used (Department of environmental affairs, 2015). the area converted is the total area converted since 1990.



5.4.7.4.1 Uncertainties

a full assessment of the uncertainties in the forest land category have not yet been completed, and will be undertaken in the next inventory. Much of the uncertainty associated with the land component relates to the mapping and area of each land type. these are discussed in Section 5.4.4. the statistics from the FSa haveahighconfidencerating(80%)withanuncertaintyrange from -11% to 3% based on a comparison with the rSa yearbook (Department of environmental affairs and tourism, 2009). Uncertainty on a lot of the activity datafortheothervegetationsub-categorieswasdifficultto estimate due to a lack of data. Uncertainty would, however, be higher than that for the forestry industry. Uncertainty of aGB growth increment in natural forests and woodlands could not be determined, but for thickets the uncertainty is 1.8 ± 0.96 t dm ha-1. the uncertainty in BCeF could not be determined; however, ipCC 2006

Guidelines suggest that it can reach up to 30%. Default root-to-shoot ratios for forests and woodlands have a range of 0.28 – 0.68; while thicket ratios had a variance of about 5%. the uncertainty on the fuel wood harvest numbers from the Fao were not provided, but it is expected to be high.

For default soil organic C stocks for mineral soils there is a nominal error estimate of ±90% (ipCC 2006 Guidelines, p 2.31). the error on the stock change factors is indicated in Table 5.13.

the uncertainty on the burnt area, MB and Cf factors is discussed in Section 5.5.2.5.

the forest land category covers a wide span of vegetation types, from closed forests to open woodlands, so the variation on the activity data was large (see Appendix F) and requires more vigorous analysis to determine the uncertainty. it is suggested that a more detailed uncertainty analysis be done on the data.

table 5.14: Sector 3 aFolU – land: relative stock change factors (± error) for different management activities on croplands (Source: Department of environmental affairs, 2015; Moeletsi et al., 2013).

Cropland FLU FMG FI

Annual crops - irrigated 0.8a (±9%) 1c (Na) 1e (Na)

Annual crops – dry 0.5g (±9%) 1c (Na) 1e (Na)

Subsistence/semi-commercial crops 0.5g (±9%) 1c (Na) 0.95f (±13%)

Orchards 0.8b,g (±50%) 1.1d (±5%) 1e (Na)

Viticulture 0.8b,g (±50%) 1.1d (±5%) 1e (Na)

a long term cultivated;

b perennial/tree crop;

c frequent tillage;

d no till;

e medium inputs;

f low inputs;

g from NtCSa (Department of environmental affairs, 2015).




the same sources of data and land-cover maps were used throughout the 12-year period so as to provide a consistent time series of data.


the land area in the land-cover maps were compared to previous data sets and explanations for the various changes have been provided. the land areas were checked and summed across all land-use categories to ensure that the total area in the inventory remained constant over the time series. Where possible, the activity data was compared to literature, previous inventories and to default values. the biomass data was compared to the biomass and Npp data from the NtCSa (Department of environmental affairs, 2015) outputs. it was noted that the inventory produces biomass growth values at the low end of the Npp range and that savannas are estimated to have a higher Npp than thickets, however as more detail is included in the next inventory so these differences are expected to be reduced. Further assessments will be completed in the next inventory. Soil carbon data was also compared to the NtCSa outputs.

as validation the plantation data was compared to a more detailed study by alembong (2015). this inventory overestimates the gains and also the losses from plantations. this could be due to a difference in plantation areas being used. However, alembong (2015) estimated the overall carbon gains in plantations in 2011 to be 857 300 t C, while the inventory estimates 558 223 t C. these values are within a similar range and more detailed will be incorporated into the next inventory to reduce these differences.

Wood removed due to harvesting were cross-checked with the input data for HWp. there was also a cross-check between the wood removed for fuelwood and the fuelwood consumption data used in the energy sector. the amount of wood harvested was compared to the Fao data. all general QC procedures discussed

in Section 1.4.1 were carried out on this category to ensure consistency in the data. an independent reviewer provided comment on the forest land category and the inventory was improved accordingly.


recalculations were made for 2000 to 2012 as new higher resolution maps were incorporated. there was also a reclassificationof thewoodland/openbushcategory.these changes impacted the area of forest land remaining forest land as well as land converted to forest land. Minor updates were made to the activity data and the DoM poolwasincluded.Landclasseswerealsostratifiedbysoil type to improve the soil carbon data.

5.4.7.7 Category-specificPlanned Improvements and Recommendations

Due to a lack of disaggregation within the land classes average growth increments for the ecosystems were taken from the literature. these could be over-estimated as there is little information on whether these growth increments are for regenerating or mature systems. it is therefore recommended that more detailed age class data be incorporated into future inventories to get a more accurate carbon gain values for the forest land.

another important recommendation is to assess and update the BCeF for fuelwood data. a reduction in uncertainty in this area would lead to improved data for the forest land category.

the next inventory plans to make use of the recently released ipCC 2006 version of the agriculture and land Use (alU) software. this software allows spatial data to be integrated with ground based data. For the forest land category the plan is to incorporate more detail regarding the forests and their age classes and management. Data from alembong (2015) will be obtained and incorporated where possible. a complete uncertainty analysis on the land data is also planned for the next inventory and the


Figure 5.16: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in Croplands between 2000 and 2012, differentiated by source category.

Figure 5.15: Sector 3 aFolU – land: trend in emissions and sinks (in Gg Co2) in Croplands between 2000 and 2012, differentiated by sub-category.



alU software will also be able to assist with this. the other advantage of the alU software is that it provides a place for all the metadata and data sources to be stored.

Going forward it is suggested that a stock-based approach be investigated for use in the inventory. this could be based on the carbon stock maps generated during the NtCSa (Department of environmental affairs, 2015) together with stock estimates for plantations and other forms of agriculture or production areas. the carbon stock estimates for indigenous landscapes could be updated on an annual or biannual time-scale using the model developed for the NtCSa. a stock-based approach allows a more complete evaluation of the change in stocks overtime,comparedtothecurrentflux/processbasedapproach that may not be complete.

5.4.8 Cropland [3b2]


reporting in the cropland category covers emissions and removals of Co2 from mineral soils, and from above- and below-ground biomass. Croplands include annual commercial crops, annual semi-commercial or subsistence crops, orchards, and viticulture. this category reports emissions and removals from the category cropland remaining cropland (cropland that remains cropland during the period covered by the report) and the land converted to cropland category. Calculations are carried out on the basis of a 23-year transition period in that once a land area is converted it remains in the converted land category for 20 years. in this inventory transition data was only available from 1990 therefore all calculations include transitions since 1990.

For Cropland remaining cropland, the tier 1 assumption is that for annual cropland there is no change in biomass carbon stocks after the f irst year (GpG-aFolU, section 5.2.1, ipCC, 2006a). the rationale is that the increase in biomass stocks in a single year is equal to the biomass losses from harvest and mortality in that

same year. For perennial cropland, there is a change in carbon stocks associated with a land-use change. Where there has been land-use change between the Cropland subcategories, carbon stock changes are reported under Cropland remaining cropland. Cropland remaining cropland was estimated to be a sink of Co2, declining from 1 241 Gg Co2 in 2000 to 104 Gg Co2 in 2012 (Figure 5.15). the sink is due to conversions between annual and perennial cropland categories and changes in soil carbon. Land converted to cropland was a source of Co2 (5 721 Gg Co2 per year) (Figure 5.16). the source remained constant as a constant annual change was derived over the mapping period between 2000 and 2013-14. it is assumed that equilibrium is reached after a year of conversion therefore even though the land converted to cropland area increased every year the change in biomass is only determined for the annual area converted. the largest contributors to the Co2 source in this category were conversions from grasslands to croplands. SoC was the largest contributor to this category (Figure 5.16). For Croplands the DoM was not estimated, but is not considered as a major contributor to this category.


5.4.8.2.1 Biomass

Cropland remaining cropland

according to the ipCC, the change in biomass is only estimated for perennial woody crops because for annual crops the increase in biomass stocks in a single year is assumed to equal the biomass losses from harvest and mortality in that same year. perennial woody crops (e.g. treecrops)accumulatebiomassforafiniteperioduntilthey are removed through harvest or reach a steady state where there is no net accumulation of carbon in biomass because growth rates have slowed and incremental gains from growth are offset by losses from natural mortality or pruning. after this period, perennial woody crops are replaced by new ones and carbon stored in biomass is


released to the atmosphere. Default annual loss rate is equal to biomass stocks at replacement. Biomass stock changes in perennials were calculated as follows:

ΔCB=A*(ΔCG–ΔC)

(eq. 5.24)

Where

ΔCB =annualchangeincarbonstocksinbiomass (tonnes C yr-1);

a = annual area of cropland (ha);

ΔCG =annualgrowthrateofperennialwoody biomass (tonnes C ha-1 yr-1);

ΔCL =annualcarbonstockinbiomassremoved (tonnes C ha-1 yr-1)

the tier 1 default assumption is that there is no change in below-ground biomass of perennial trees or crops in agricultural systems (ipCC 2006 Guidelines, p. 5.10). there are no default values and so below-ground biomass was assumed to be zero.

the Co2 emissions from biomass burning in annual Croplands does not have to be reported, as the carbon released during combustion is assumed to be reabsorbed by the vegetation during the next growing season. Co2 emissions from the burning of perennial crops were included by using equation 2.27 from ipCC 2006 Guidelines (see Section 5.5.2).

Land converted to croplands

For this a tier 1 combined with a tier 2 approach was applied. the annual increase in carbon stocks in biomass due to land conversions was estimated using the following ipCC 2006 equation:

ΔCB=ΔCG + ((BaFter-BBeFore)*ΔAto_otHer)*CF–ΔCl

(eq. 5.25)

Where:

ΔCB = annual change in carbon stocks in biomass (t C yr-1);

ΔCG = annual biomass carbon growth (t C ha-1 yr-1);

BaFter = biomass stocks on the land type immediately after conversion (t dm ha-1);

BBeFore = biomass stocks before the conversion (t dm ha-1);

ΔAto_otHer = annual area of land converted to cropland (ha);

CF = carbon fraction of dry matter (t C/t dm-1);

ΔCl = annual loss of biomass carbon (t C ha-1 yr-1).


Cropland remaining cropland and Land converted to cropland

Soil carbon sources and sinks were calculated as described in Section 5.4.7.2.3. the land areas for the various categories were obtained from the land-use change maps (Geoterraimage, 2015) and described in Section 5.4.4. For the croplands remaining croplands category this included conversions between croplands, and for the land converted to croplands category the land area included all converted land since 1990.




5.4.8.3.1 Biomass

Croplands remaining croplands

the areas of the various croplands remaining croplands and land converted to croplands were obtained from the updated land-use maps developed by Geoterraimage (2015) and are discussed in detail in Section 5.4.4. For croplands remaining croplands the area is multiplied by a carbon gain and loss. Default net accumulation rates and biomass stocks were considered for perennial crops, however according to data from the NtCSa (Department of environmental affairs, 2015) these values are high. the Dea (2015) report provided biomass carbon for cultivated land in general, therefore to obtain a value for orchards and vineyards the above-ground carbon (woody and herbaceous) layers from this study were integrated into arcGiS 10 and extrapolated based on the 2013/14 land-cover dataset. this provided a value of 59 t dm ha-1 yr-1 for orchards and a value of 22 t dm ha-1 yr-1 for vineyards (see Appendix F). there is not a lot of data available for validation but the National inventory report for Greece has a value of 48 – 54 t dm ha-1 for orchards and for vineyards biomass data of 14 – 17.8 t dm ha-1 are reported in the USa (Morande, 2015). these suggest that the calculated aGB from the Dea (2015) report are reasonable and were therefore applied in this inventory. Considering statistics for orchards and vineyards (Hortgro, 2015) the age distribution of the crops is shown to be up to 18 years plus and 25 years plus for various orchard types and up to 25 years plus for vineyards. Based on this it was assumed that on average the orchards and vines grow for 25 years. this is a broad assumption, and may be an underestimate, as the category is for orchards in general. More cropland detail and age class distribution will be included in the next inventory so these assumptions and estimates can be improved on. Biomass was assumed to accumulate linearly for the entire 25 year period, therefore, the growth rate was calculated as the biomass divided by age. these derived growth

rates (1.16 t dm ha-1 for orchards and 0.44 t dm ha-1 for vineyards) are much lower than the ipCC default values, but similarly low growth rates have been used by other countries (National inventory report, New Zealand). Forlossestheareaharvestedwasdifficulttodetermine.the South african Wine industry Statistics report (2015) provides the area of uprooted vineyards and this was taken to be the area for loss. the biomass stock was multiplied by this area to calculate the carbon loss for vineyards. the activity data available for orchards does not provide information on areas of perennial cropland temporarily destocked; therefore, no losses in carbon stock due to temporary destocking could be calculated at this stage and can be considered in future inventories if data becomes available.

Land converted to cropland

it is assumed that land is cleared before it is converted to a crop, therefore the biomass stock immediately after conversion is assumed to be zero. indigenous forest aGB are provided in section 5.4.7.3.1. For thickets and dense bush the 60 t dm ha-1 (Mills et al., 2005; Mills and Cowling, 2006; lechmere-oertel, 2003) used in the previous inventory was applied. this was within the range provided by the NtCSa (Department of environmental affairs, 2015). this study reported an average value of 50 t dm ha-

1, and if this carbon data is overlaid with the updated 2014 land cover map then a value of 68 t dm ha-1 is obtained (see Appendix F). therefore, the value identified from the literature is within this range. the woodland/dense bush aGB values were adjusted from the previous inventory due to data from the NtCSa (Department of environmental affairs, 2015), but also due to changes in classification.Thisvegetationclassisverylargeandcoverssavannas and woodlands so the variation and uncertainty on the average value is large. intersecting the new land cover map (2014) with the woody and herbaceous carbon data from the Dea (2015) report provided a value of 14.86 t dm ha-1. this is within the range of 3.4 – 57.8 t dm ha-1 sourced from the literature (Scholes and Hall, 1996; rutherford, 1982; Shackelton and Scholes, 2011;


Colgan et al., 2012). the average aGB value of 18 t dm ha-1 provided by Scholes and Hall (1996) was used in this inventory. the biomass data for other land classes before conversions are provided in Table 5.15. the ipCC default values for annual cropland and grassland aGB were used as in the previous inventory as the data from the NtCSa (Department of environmental affairs, 2015) are in the same range (see Appendix F).

an ipCC default carbon gain value of 5 t C ha-1 was applied for annual croplands (ipCC 2006, vol 4, chapter 5, page 28, table 5.9). Gains and losses for phase 2 of the equation were determined as for cropland remaining cropland. the carbon fraction was the default of 0.5 t C/t dm-1.


Data are as described in Section 5.4.7.3.3.



the main uncertainties were associated with the area estimates of the croplands and this is discussed in Section 5.4.4. Uncertainty on the biomass carbon stocks for forest land categories was not determined, but default biomass carbon stocks have an error of ±75%. ipCC default values for carbon stocks after one year of growth in crops planted after conversion also have an error of ±75% (ipCC, 2006, p. 5.28).

table 5.15: Sector 3 aFolU – land: above ground biomass for the various land classes.

Land classbiomass carbon stock

(t dm ha-1)Data source

plantations eucalyptus grandis other eucs Softwood pulp acacia other sp.

841117113292

values for plantation species were calculated from FSa data and Dovey and Smith (2005) dry matter ratios.

thicket 60 previous inventory

Woodland/open bush 18

Scholes and Hall, 1996; rutherford, 1982; Shackelton and Scholes, 2011; Colgaren et al., 2012. also supported by the outputs of the intersection of Dea (2015) aGB data and the new 2014 land cover map

Grasslands 6.1 ipCC 2006 Guidelines

Cultivated annual 10 Dea (2015); ipCC 2006 Guidelines

orchards 59 overlay of Dea (2015) aGB data and the new 2014 land cover mapvines 22



For default soil organic C stocks for mineral soils there is a nominal error estimate of ±90% (ipCC 2006 Guidelines, p 2.31). the error on the stock change factors is indicated in Table 5.14. the biomass and DoM activity data is highly variable and the standard deviation on these data sets are shown in Appendix F.


the same sources of data and land cover maps were used throughout the 12 year period so as to provide a consistent time-series of data.


the land area in the land-cover maps was compared to other data sets. land-use change and activity data was compared to the literature and other data sources, where possible. the land areas were checked and summed across all land-use categories to ensure that the total area in the inventory remained constant over the time series. all general QC procedures discussed in Section 1.4.1 were carried out on this category to ensure consistency in the data. an independent reviewer provided comment on the crop land category and the inventory was improved accordingly.


the updated land cover maps meant that recalculations had to be done for all years between 2000 and 2012. Updated activity and soil carbon data were also incorporated.


Under the GHG improvement programme initiated by the Dea, a project was recently completed by the agricultural research Council (arC) which provides more detailed cropland data. Due to the timing of the project the data could not be incorporated into this inventory but will be considered in the next one. the

next inventory will make use of the alU software which will enable a more detailed categorization of croplands, and also allow for the integration of the detailed crop management data (irrigation, fertilization, rotations). the improved management data will also assist in improving the soil carbon data. Detailed categorization of crops will alsoallowmorecropspecificbiomassandgrowthdatato be included which can aid in reducing uncertainty. it is important for South africa to move towards a tier 2 for croplands.

5.4.9 Grassland [3b3]


the Grassland remaining grassland category includes all grasslands, managed pastures and rangelands. the ipCC does recommend separating out improved grassland; however, there was insufficient information at the national scale to enable this division so all grasslands wereclassifiedtogether.Thissectiondealswithemissionsand removals of Co2 in the biomass and mineral soil carbonpools,buttherewasinsufficientdatatoincludethe DoM pool. Co2 emissions from biomass burning were not reported since they are largely balanced by the Co2 that is reincorporated back into the biomass via photosynthetic activity.

Grasslands remaining grasslands were assumed to be in balance so there were no emissions from this sub-category. Land converted to grasslands were estimated to be a source of Co2 (7 904 Gg Co2 per year) because of land use conversions. the sink remained constant, as a constant annual change was derived over the mapping period between 2000 and 2013-14. it is assumed that equilibrium is reached after a year of conversion, therefore, even though the land converted to grassland area increased every year, the change is only determined for the annual area converted. DoM was not estimated for grasslands.



5.4.9.2.1 Biomass

Grassland remaining grassland

a tier 1 approach assumes no change in biomass in grassland remaining grassland, as carbon accumulation through plant growth is roughly balanced by losses throughgrazing,decompositionandfire.

Land converted to grassland

annual change in biomass carbon stocks on land converted to grasslands was estimated using equation 5.25 above. the average carbon stock change is equal to the carbon stock change due to the removal of biomass from the initial land use plus carbon stocks for biomass growth. tier 1 method assumes that biomass is cleared when preparing a site for grassland use, so the default for biomass immediately after conversion is zero. the other assumption is that grasslands achieve their steady statebiomassduringthefirstyearfollowingconversionso there are no stock changes associated with phase 2. the carbon fraction was the default of 0.5 t C /t dm-1.


Grasslands remaining grasslands and Land converted to grasslands

Soil carbon sources and sinks were calculated as described in Section 5.4.7.2.2. the land areas for the various categories were obtained from the land-use change maps (Geoterraimage, 2015). For land converted to grasslands the land area includes all converted land since 1990.


5.4.9.3.1 Biomass

the areas of grassland remaining grassland and annual area of land converted to grassland were obtained from the updated land use maps developed by Geoterraimage (2015) and are discussed in detail in section 5.4.4. the BBeFore aGB values are provided in Table 5.15 and in the relevant biomass sections.


Data are as described in Section 5.4.7.3.3.



the main uncertainties were associated with the area estimates of the grassland categories (discussed in Section 5.4.4). Grasslands achieved a user accuracy of 85% but a producer’s accuracy of 70%, which resulted from the confusion between grasslands, woodland/open bush and low shrubland (Gti 2013). a full uncertainty assessment has not been conducted, but is this will be done in the next inventory. Uncertainty on the biomass carbon stocks for the various land classes is given in Appendix F. Default biomass carbon stocks for the various land categories have an error of ±75%. ipCC default values for carbon stocks after one year of growth in crops planted after conversion also have an error of ±75% (ipCC, 2006, p. 5.28).

For default soil organic C stocks for mineral soils there is a nominal error estimate of ±90% (ipCC 2006 Guidelines, p 2.31). the error on the stock change factors is indicated in Table 5.13.




the same sources of data and land-cover maps were used throughout the 12-year period to provide a consistent time series of data.


the land area in the land-cover maps was compared to other data sets. land-use change and activity data were compared to the literature and other data sources where possible;otherwisenoothersource-specificQA/QCprocedures were carried out on this subcategory. the land areas were checked and summed across all land-use categories to ensure that the total area in the inventory remained constant over the time series. all general QC procedures discussed in Section 1.4.1 were carried out on this category to ensure consistency in the data. an independent reviewer provided comment on the forest land category and the inventory was improved accordingly.


Due to the updated land-cover maps and improved soils data recalculations were done for all years between 2000 and 2012.


Nospecificimprovementsareplannedforthiscategory,however it is recommended that in future a more detailed classificationofgrasslandsbeincludedsoastoimprovethe soil carbon for this category and to reduce uncertainty.

5.4.10 Wetlands [3b4]


Waterbodies and wetlands are the two sub-divisions in thewetlandcategoryandthesearedefinedinsection 5.4.5.5. peatlands are included under wetlands, and due to the resolution of the mapping approach used, the area of peatlands could not be distinguished from the other wetlands, therefore they were grouped together. Co2 emissions from wetlands remaining wetlands and land converted to wetlands are not reported, however an estimate is provided for CH4fromfloodedlands.

in 2000 wetlands produced 429 Gg Co2eq and this declined by 37.7% to 267 Gg Co2eq in 2012 (Figure 5.17). Thisdeclineisrelatedtothereducedareaoffloodedlands as the wetland area decreased by 9.2% between 2000 and 20125.


CH4 emissions from wetlands were calculated as in the previous inventory following the equation:

CH4 emissionsWWFlood = p * e(CH4)diff * a * 10-6

(eq. 5.26)

Where:

CH4 emissions WWFlood = total CH4 emissions from floodedland(GgCH4 yr-1);

p = ice-free period (days yr-1);

e(CH4)diff = average daily diffusive emissions (kg CH4 ha-1 day-1);

A =areaoffloodedland(ha).

5 it should be noted that this decreased wetland area should be considered with some caution as 1990 was a wetter year than 2013-14 so this reduction may be overestimated.



the area of waterbodies was taken from the Geoterraimage (2014) land use maps. For South africa the ice-free period is taken as 365 days. the emission factor (e(CH4)diff) was selected to be a median average for the warm temperate dry climate values provided in Table 3.A2 (ipCC 2006, volume 3). this emission factor is the lowest of all climates and therefore provides a conservative estimate.


the main uncertainties were associated with the area estimates of the wetland categories (discussed in Section

5.4.4). the same sources of data and land-cover maps were used throughout the 12-year period to provide a consistent time series of data.

5.4.10.5Source-specificRecalculations

the 2000 to 2012 emissions were recalculated using the updated wetland area.

5.4.10.6Source-specificPlannedImprovements and Recommendations

No improvements are planned for this category.

Figure 5.17: Sector 3 aFolU – land: trend in CH4 emissions (in Gg Co2eq) in wetlands between 2000 and 2012.



5.4.11 Settlements [3b5]


Settlements include all formal built-up areas, in which people reside on a permanent or near-permanent basis. it includes transportation infrastructure as well as mines. the population of South africa in 2010 was just under 50 million (Statistics South africa, 2010), with the urban population increasing from 52% to 62% over the past two decades (Department of environmental affairs, 2011). the surface area of settlements provided by the land cover map developed for this inventory was 3 228 445 ha (2 901 090 ha of settlements and 327 354 ha under mines) in 2012. this area was higher than the area used in the 2010 inventory. Changes in the extent of urban areas between 1990 and 2013-14 (increase of 6.7%) may not beaslocallysignificantasexpectedasthesettlementscategory includes peripheral smallholding areas around themainbuilt-upareas;andthesetendtobethefirstland-use that is converted to formal urban areas, before further expansion into natural and cultivated lands.

this category assumes there are no emissions from settlements remaining settlements, but includes emissions and sinks from biomass and soil carbon pools for land converted to settlements. this category was found to be a source of Co2 (1 195 Gg Co2) due to land conversions. the soil source is approximately a third of the carbon, with the rest being from biomass.


5.4.11.2.1 Biomass

Land converted to settlements

annual change in biomass carbon stocks on land converted to settlements was estimated using equation 5.25 above.

a tier 1 method assumes that all biomass is cleared when preparing a site for settlements, thus the default for biomass immediately after conversion is zero. above-ground biomass of the various land categories was taken from Table 5.15. as the biomass value before conversion. it is also assumed that there is no additional growth or loss of carbon on the land after it becomes a settlement. the carbon fraction was the default of 0.5 t C/t dm-1.


Land converted to settlements

Soil carbon sources and sinks were calculated as described in Section 5.4.7.2.3.


5.4.11.3.1 Biomass

the area of land converted to settlements was obtained from the updated land use maps developed by Geoterraimage (2015) and are discussed in detail in section 5.4.4. Biomass carbon stocks before conversion for the various vegetation types are provided in Table 5.15. and the relevant biomass sections.


the land areas for the various categories were obtained from the land-use change maps (Geoterraimage, 2015) and described in Section 5.4.4. For land converted to settlements the land area includes all converted land since 1990. all other data are described in Section 5.4.7.3.3. a stock change factor of 0.8 was applied to settlements.




the main uncertainties were associated with the area estimates of settlements and this is discussed in Section 5.4.4.


the same sources of data and land-cover maps were used throughout the 12-year period so as to provide a consistent time series of data.

5.4.11.5Source-specificQA/QCandVerification

the land area in the land-cover maps was compared to other data sets. land-use change and activity data were compared to the literature and other data sources, where possible,otherwisenoothersource-specificQA/QCprocedures were carried out on this subcategory.


Due to new land areas from the updated land-cover maps, recalculations were done for all years between 2000 and 2012.

5.4.11.7Category-specificPlanned Improvements and Recommendations

there are no planned improvements for this category, particularly since it is not a key category. However carbon stock changes could be improved if land conversions and country-specificdataonbiomassinsettlementscouldbe included.

5.4.12 Other land [3b6]


other land includes bare soil, rock, and all other land areas that do not fall into the other land classes (including low shrublands). this category includes emissions and sinks for other land remaining other land and land converted to other lands. there are assumed to be no changes in the Other land remaining Other land category, although this should be investigated further in future submissions as there can be a change in carbon if low shrubland is converted to bare ground for example. For the land converted to other land category both the biomass and soil carbon changes are included. this category was found to have an annual sink of 960 Gg Co2. Soils were the largest contributors to this category although biomass has not been fully incorporated.


5.4.12.2.1 Biomass

Land converted to other lands

annual change in biomass carbon stocks on land converted to other lands was estimated using equation 5.25 above. Carbon stocks before the conversion are provided in Table 5.15. and in the relevant biomass sections. Following the tier 1 approach, it was assumed that other lands achieved their steady-state biomass duringthefirstyearfollowingconversionsonostockchanges were associated with phase 2 (i.e. no change due to growth or losses). the carbon fraction was the default of 0.5 t C/t dm-1.




Land converted to other lands

Soil carbon sources and sinks were calculated as described in Section 5.4.7.2.3. according to ipCC tier 1 method the reference C stock at the end of the 23 year transition period is assumed to be zero. this, however, only applies to areas that have no vegetation or have no possibility of havingvegetationonthem.SouthAfrica’sclassificationof Other land includes low shrublands, bare ground and degraded land all of which have some form of vegetation on them (see Gti report in Appendix G). only a small portion of the bare ground category has no vegetation or does not have the potential to have vegetation on it. Based on this the assumption of zero carbon after 23 years was not assumed to be the case and the ipCC default SoC reference values were applied as in the other land categories.


5.4.12.3.1 Biomass

the area of land converted to settlements was obtained from the updated land use maps developed by Geoterraimage (2015) and are discussed in detail in section 5.4.4. Biomass carbon stocks before conversion for the various vegetation types are shown in Table 5.15. and in the relevant biomass sections.


the land areas for the various categories was obtained from the land-use change maps (Geoterraimage, 2015) and described in Section 5.4.4. For land converted to other lands the land area includes all converted land since 1990. all other data are described in Section 5.4.7.3.2.



the main uncertainties were associated with the area estimates of other lands and this is discussed in Section 5.4.4. Standard deviations on biomass activity data are provided in Appendix F.


the same sources of data and land-cover maps were used throughout the 12-year period to provide a consistent time series of data.

5.4.12.5Source-specificQA/QCandVerification

the land area in the land-cover maps was compared to otherdatasets.Nosource-specificQA/QCprocedureswere carried out on this sub-category.


recalculations were completed for all years between 2000 and 2012 due to updated land areas.

5.4.12.7Category-specificPlanned Improvements and Recommendations

Changes in biomass for other land remaining other land will be included in the next inventory. other than that no other improvements are planned for this category.


5.5 Aggregated sources and non- CO2 emission sources on land [3C]


aggregated and non-Co2 emission sources on land produced an accumulated total of 302 275 Gg Co2eq between2000and2012.Theannualemissionsfluctuatedbetween a low of 22 529 Gg Co2eq in 2012 and a high of 24 607 Gg Co2eq in 2002. there was a lot of annual variation in emissions from each of the sub-categories in this section, with none of them showing a clear increasing or decreasing trend (Figure 5.18). Direct N2o emissions

from managed soil was the biggest contributor to this category, producing between 65.3% (2012) and 67.6% (2000) of the total annual aggregated and non-Co2 emissions. this was followed by indirect N2o emissions from managed soils (20.3% in 2000 to 20.2% in 2012) and biomass burning (8.1% to 7.5 %).

5.5.2 biomass burning [3C1]


Biomass burning is an important ecosystem process in SouthernAfrica,withsignificantimplicationsforregionaland global atmospheric chemistry and biogeochemical cycles (Korontzi et al., 2003). according to the National

Figure 5.18: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission levels, 2000 – 2012. MM = Manure management; MS = Managed soils; BB = Biomass burning.



inventory report (Department of environmental affairs andTourism,2009),fireplaysanimportantroleinSouthafrican biomes, where grassland, savanna and Fynbos firesmaintainecologicalhealth.Inadditiontocarbondioxide, the burning of biomass results in the release of other GHGs or precursors of GHGs that originate from incomplete combustion of the fuel. the key GHGs are Co2, CH4, and N2o; however, Nox, NH3, NMvoC and Co are also produced and these are precursors for the formation of GHG in the atmosphere (ipCC, 2006).

although the ipCC Guidelines only require the calculation of emissions from savanna burning, South africa reports emissions of non-Co2 gases (CH4, Co, N2o and Nox) from all land categories, as explained in the 2000 inventory (Department of environmental affairs and tourism, 2009). Theburningofbiomassisclassifiedintothesixland-use

categoriesdefinedinthe2006Guidelines,namely,forestland, cropland, grassland, wetlands, settlements and other land. the ipCC Guidelines suggest that emissions from savanna burning should be included under the grassland category; however, since, in this inventory woodlands andopenbushhavebeenclassifiedasforestland,theiremissions were dealt with under forest land.

although the burning of croplands might be limited, burning has been shown to occur on cultivated land (archibald et al.,2010),mainlyduetothespreadoffiresfrom surrounding grassland areas. the croplands category was sub-divided into annual and perennial crops; there was also a sub-division for sugarcane as residue burning is still an acceptable practice in South africa (Hurly et al., 2003).

Figure 5.19: Sector 3 aFolU – aggregated and non-Co2 sources: Contribution of the various land categories to the biomass burning emissions, 2000 – 2012.


the Co2 net emissions should be reported when Co2

emissions and removals from the biomass pool are not equivalent in the inventory year. For grasslands and annual croplands the annual Co2 removals (through growth) and emissions(whetherbydecayorfire)areinbalance.CO2 emissions are therefore assumed to be zero for these categories.

Non-Co2 emissions from biomass burning in all land categories were dealt with in this section. For forest land the Co2 emissions from biomass burning were not included in this section but rather in the forest land section (see Section 5.4.7).


Biomass burning contributed between 1 576 Gg Co2eq (2004) and 2 419 Gg Co2eq (2005) to the overall emissions in the aFolU sector. of this, about 53% was CH4 and 47% N2o. Grasslands contributed the most to biomass-burning emissions (62.0% to 54.7 %), followed by forest lands (17.1% to 17.4 %) and croplands (16.6% to 20.33 %) (Figure 5.19). in the previous inventory (Department of environmental affairs, 2014), grasslands contributed about 52% and forest lands 28% in 2010, compared with the 55.1% and 16.3%, respectively, in 2010 in this inventory. this change was likely due to changes in land area that were brought about by the incorporation of the new land-cover maps, as well as the updated emission factors.


the tier 2 methodology was applied, with the emissions from biomass burning being calculated using the following equation (equation 2.27 from ipCC 2006 Guidelines):

lfire = a * MB * Cf * Gef * 10-3

(eq. 5.27)

Where:

lfire =massofGHGemissionsfromthefire(tGHG)

a = area burnt (ha)

MB = mass of fuel available for combustion (t dm ha-1)

Cf = combustion factor (dimensionless)

Gef = emission factor (g kg-1 dm burnt)


5.5.2.4.1 Burnt area data

annual burnt-area maps were produced from the MoDiS monthly burnt-area product for each year of the inventory (2000 to 2012). the MoDiS Collection 5 Burned area product (MCD45) Geotiff version from the University of Maryland (ftp://ba1.geog.umd.edu) was used. this is a level 3 gridded 500 m product and the quality of the information is described in Boschetti et al. (2012). every month of data was reprojected into the UtM 35S projection to remain consistent with the 2013-14 land-cover dataset project. all the monthly maps were then merged into an annual map by adding the valid burnt areas in each map. this was done for each year between 2000 and 2012. these burnt-area maps were then overlaid with the 2013-14 land-cover maps to determine the burnt area in each land class. the percentage of burnt area in each land class is shown in Table 5.16.

Some burnt area corrections were applied to the data. it was assumed that there was no burning in indigenous forests and thickets, so any burnt area under these categories was allocated to grasslands. the FSa reports burntareaforplantationssothesefigureswereusedandany difference between the FSa and MoDiS burnt area was added or subtracted from grasslands. any burning on bare ground was allocated to low shrublands, and any burning under water bodies was allocated to wetlands.



tabl

e 5.

16:

Sect

or 3

aFo

lU –

agg

rega

ted

and

non-

Co

2 sou

rces

: per

cent

age

area

bur

nt w

ithi

n ea

ch la

nd c

lass

bet

wee

n 20

00 a

nd 2

012.

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

Indi

geno

us fo

rest

0.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

00

Thi

cket

0.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

00

Woo

dlan

d / O

pen

bush

5.88

8.46

6.54

2.51

4.69

7.92

6.87

3.95

7.14

4.71

8.09

6.94

6.97

Pla

ntat

ions

1.08

0.92

0.89

1.55

1.51

1.20

1.54

3.77

3.78

1.06

0.84

0.82

0.51

Ann

ual c

rops

: piv

ot3.

913.

053.

963.

674.

625.

083.

813.

822.

702.

804.

183.

192.

32

Ann

ual c

rops

: non

-piv

ot3.

883.

594.

403.

282.

774.

694.

782.

703.

133.

093.

904.

312.

70

Orc

hard

0.

601.

150.

840.

860.

471.

060.

741.

281.

640.

701.

460.

971.

34

Vit

icul

ture

0.

540.

350.

280.

440.

110.

310.

610.

290.

220.

330.

120.

290.

01

Subs

iste

nce

crop

s

5.44

8.14

6.41

5.44

4.34

9.40

7.37

11.4

16.

868.

129.

885.

984.

62

Sett

lem

ents

1.94

2.15

1.90

1.64

1.28

2.90

2.54

2.81

1.73

2.08

2.76

1.52

1.83

Min

es3.

302.

974.

262.

903.

314.

854.

712.

312.

421.

963.

092.

712.

01

Wet

land

s

7.37

8.90

10.0

78.

247.

0610

.45

9.28

8.17

7.45

8.37

8.41

8.32

5.72

Gra

ssla

nds

10.8

912

.70

12.6

99.

848.

4913

.44

11.6

211

.34

10.7

310

.65

12.1

211

.32

9.14

Wat

er b

odie

s

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Low

shr

ubla

nds

0.36

0.30

0.73

0.46

0.28

0.48

0.37

0.32

0.38

0.51

0.88

0.86

0.44

Deg

rade

d7.

196.

518.

722.

567.

709.

4313

.98

2.13

9.95

6.46

9.91

10.4

73.

35


5.5.2.4.2Massoffuelavailableforcombustion(MB)and thecombustionfactor(Cf)

the values for fuel density were sourced from various sources (Table 5.17). a weighted average for fuel density and the combustion factor (Cf) was determined for low shrublands. according to the 2013/2014 land cover map report (Gti, 2015) low shrublands are mainly Karoo type vegetation. also included in this category is a portion of Fynbos (13% according to the 2013/2014 land cover map). the karoo vegetation classes have similar fuel densities and Cf values, but these are very different for Fynbos (Table 5.17). a weighted average fuel density and Cf value was calculated from these numbers for the low shrubland category in this inventory. Wetlands were assumed to have the same values as grasslands as done

in the earlier inventories (Department of environmental affairs, 2009; 1994).

Comparing the data to ipCC values highlights a few discrepancies. the woodland/open bush and the grassland combustion factors are higher than the values provided by ipCC but this estimate is based on actual data for South africa therefore assumed to be more appropriate. the low shrubland weighted average fuel density is lower than the general shrubland values provided in ipCC. the reason for this is that for South africa this category includes arid shrublands which have much lower fuel density than the shrublands used to determine the ipCC default table (ipCC, 2006, table 2.4, vol 4, chapter 2, page 2.46).

table 5.17: Sector 3 aFolU – aggregated and non-Co2 sources: Fuel density and combustion fractions for the various vegetation classes.

Vegetation class

Fuel density (t/ha)

SourceIPCC default (Table 2.4, vol

4, chapt 2)

Com-bustion fraction

SourceIPCC value

plantations 302000 Nir (Dea, 2009)

0.42000 Nir (Dea, 2009)

Woodlands/open bush

4Hely et al. (2003)

2.6 – 4.6 0.88Hely et al. (2003)

0.4 – 0.74

Croplands 7 DaFF (2010)4 – 10 (agricul-tural residue)

1DaFF (2010)

0.8 – 0.9 (agricultural

residue)

Grasslands 4Hely et al. (2003)

2.1 - 10 0.88Hely et al. (2003)

0.74 – 0.77

low shrublands in general

2.42aWeighted average

5.7 – 26.7 0.91aWeighted average

0.61 – 0.95

Fynbos 12.9 ipCC 2006 0.61 ipCC 2006

Nama karoo 1 1994 Nir (??) 0.951994 Nir (??)

Succulent karoo 0.6 1994 Nir (??) 0.951994 Nir (??)

Wetlands 42000 Nir (Dea, 2009)

0.882000 Nir (Dea, 2009)

a See text for explanation.




ipCC 2006 default emission factors (ipCC, 2006, vol 4, chapter 2, table 2.5, page 2.47) were applied as shown in Table 5.18.



the MoDiS burnt area products have been shown to identify about 75% of the burnt area in Southern africa (roy and Boschetti, 2009). the MCD45 product produces afinerresolution(500m)thantheotherproducts(1km)and uses a more sophisticated change-detection process to identify a burn scar (roy et al., 2005). it also provides ancillary data on the quality of the burn-scar detection. the MCD45 product has been shown to have the lowest omission and commission errors compared to the l3JrC

and GlobCarbon products (anaya and Chuvieco, 2012). Much of the uncertainty lies with the land-cover maps (Section 5.4.4)andsomecorrectionsformisclassifiedpixels were made. Since the values for fuel density, combustion fraction and emission factors were taken from the 2000 GHG inventory, the uncertainties discussed in Section 5.5.1.6 of the 2000 inventory (Department of environmental affairs and tourism, 2009) still apply to this data set.


the MoDiS burnt area product was used for all 12 years to maintain consistency.

5.5.2.6 Source-specificQA/QC

the burnt area data derived in this inventory were compared to the data from the previous inventories (Department of environmental affairs and tourism, 2009; Department of environmental affairs, 2014) using the data fromthesameyears.Directcomparisonsaredifficult,however,duetothechangesintheland-classdefinitions.

in terms of the amount of fuel burnt and the combustion factors, South africa is one of the leaders in research on biomass burning. the methods used to derive the data were very comprehensive and locally relevant.

all general QC procedures to check data, transcriptions and spreadsheets, as detailed in Section 1.4.1, were also conducted.



recalculations were done for all years between 2000 and 2012 due to the introduction of new land cover data and

table 5.18: Sector 3 aFolU – aggregated and non-Co2 sources: emission factors applied to the various vegetation classes.

Vegetation class EF ± SE

plantations

Co2 1569 131

Co 107 37

CH4 4.7 1.9

N2o 0.26 0.07

Nox 3.0 1.4

Woodland/open bush;grasslands; low shrublands; wetlands

Co2 1613 95

Co 65 20

CH4 2.3 0.9

N2o 0.21 0.1

Nox 3.9 2.4

Croplands

Co2 1515 177

Co 92 84

CH4 2.7

N2o 0.07

Nox 2.5 1.0


the incorporation of updated emission factors.


Therearenospecificplanned improvements forthiscategory; however, there are plans to develop higher-resolution land-use maps. these could be overlaid with the appropriate burnt-area data to improve estimates for those years. Higher-resolution land-use maps will reduce the need for corrections.

5.5.3 Liming and urea application [3C2 and 3C3]


liming is used to reduce soil acidity and improve plant

growth in managed systems. adding carbonates to soils in the form of lime (limestone or dolomite) leads to Co2 emissions as the carbonate limes dissolve and release bicarbonate.

adding urea to soils during fertilization leads to a loss of Co2thatwasfixedintheindustrialproductionprocess.Similar to the soil reaction following the addition of lime, bicarbonate that is formed evolves into Co2 and water.


the Co2 from liming showed a high annual variability, with the highest emissions being in 2002 (684 Gg Co2) and the lowest in 2005 (267 Gg Co2) (Figure 5.20). the variation was directly linked to limestone and dolomite consumption (Figure 5.20). the Co2 emissions from urea application showed a similar annual variability (Figure

Figure 5.20: Sector 3 aFolU – aggregated and non-Co2 sources: trends and emission levels from liming and urea application, 2000 – 2012.



5.21). Urea application produced an accumulated amount of 5 156 Gg Co2 between 2000 and 2012, with the lowest emissions (147 Gg Co2) occurring in 2001 and the highest (555 Gg Co2) in 2012. there was a sharp decline in emissions from both liming and urea application in 2005.

5.5.3.3 Methodologocal Issues

a tier 1 approach of the ipCC 2006 Guidelines was used to calculate annual C emissions from lime application (equation 11.12, ipCC 2006) and Co2 emissions from urea fertilization (equation 11.13, ipCC 2006).


the amount of limestone and dolomite applied was obtained from the Fertilizer association of South africa (http://www.fssa.org.za/Statistics.html) (Table 5. 19). Data for 2011 to 2012 were not available due

to restrictions placed by the Competition Commission. annual consumption is highly variable and does not appear to have any relationship with surrogate data, making extrapolationdifficult.Sointhisinventorythevaluefrom2010 was used for all three years, but this issue needs to be addressed in the next inventory. in the previous inventory the amount of urea used was assumed to be the amount imported into South africa (FaoStat; http://faostat.fao.org/site/575/default.aspx#ancor; accessed on 02/2013). However, in this inventory the consumption was assumed to be the difference between urea imports and exports in the Fao database.


the ipCC default emission factors of 0.12 t C (t limestone)-1; 0.13 t C (t dolomite)-1; and 0.2 t C (t urea)-1 were used to calculate the Co2 emissions.

Figure 5.21: Sector 3 aFolU – aggregated and non-Co2 sources: annual amount of lime and urea applied to soils, 2000 – 2012.




the dolomite and limestone default emission factors have an uncertainty of -50% (ipCC 2006 Guidelines, p. 11.27). in terms of urea application it was assumed that all urea imported was applied to agricultural soils and this approach may lead to an over- or under-estimate if the total imported is not applied in that particular year. However, over the long-term this bias should be negligible (ipCC, 2006). as for the liming emission factors, the urea emission factor also has an uncertainty of -50% (ipCC 2006 Guidelines, p. 11.32).


For liming the same source of activity data was used for 2000 to 2009, and for 2010 numbers were estimated

based on this data so as to try and maintain the time-series consistency. the same activity data source for urea was applied to all 12 years.


Quality control of the activity data was limited for this section as very little comparative data was available. Data, calculations and spreadsheets were all checked to ensure data was carried through properly and that there were no transcription errors. an external reviewer provided the quality assurance.


emission data for urea were recalculated for all years between 2000 and 2012 due to the adjustment made to the urea consumption data.


at this stage there are no improvements planned for Co2 emissions from liming and urea application. However, more accurate urea application data would improve the emission estimates.

5.5.4 Direct N2O emissions from managed soils [3C4]


agricultural soils contribute to GHGs in three ways (Desjardins et al., 1993):

• Co2 through the loss of soil organic matter. this is a result of land-use change, and is, therefore, dealt with in the land sector, not in this section;

• CH4 from anaerobic soils. anaerobic cultivation, such as rice paddies, is not practised in South africa, and therefore CH4 emissions from agricultural soils are

table 3.19: Sector 1 energy: Fugitive emissions from coal mining for the period 2000 to 2012.

YearAmount used (t yr-1)

limestone Dolomite Urea

2000 254 116 571 136 288 400

2001 329 996 738 361 200 000

2002 436 743 1 031 172 670 700

2003 473 006 792 736 451 636

2004 474 215 790 673 578 942

2005 253 606 326 898 463 086

2006 357 970 605 148 472 523

2007 474 753 662 893 627 196

2008 616 844 812 959 607 729

2009 508 526 823 382 501 526

2010 487 816 667 564 660 707

2011 487 816 667 564 751 527

2012 487 816 667 564 756 827



not included in this inventory; and

• N2o from fertilizer use and intensive cultivation. Thisisasignificantfractionofnon-carbonemissionsfrom agriculture and is the focus of this section of the inventory.

TheIPCC(2006)identifiesseveralpathwaysofnitrogeninputs to agricultural soils that can result in direct N2o emissions:

• Nitrogen inputs:

o Synthetic nitrogen fertilizers;

o organic fertilizers (including animal manure, compost and sewage sludge); and

o Cropresidue(includingnitrogenfixingcrops);

• Soil organic matter lost from mineral soils through land-use change (dealt with under the land sector);

• organic soil that is drained or managed for agricultural purposes (also dealt with under the land sector); and

• animal manure deposited on pastures, rangelands and paddocks.


Direct N2o emissions from managed soils produced an accumulated total of 198 528 Gg Co2eq between 2000 and 2012.Emissionsfluctuatedannuallywith2002havingthehighest (15 909 Gg Co2eq and 2012 the lowest emissions of 14 701 Gg Co2eq (Figure 5.22). the variation was duemainlytothefluctuationinmanureandsyntheticfertilizer inputs, while emissions from compost, sewage sludgeandcropresiduesdidnotchangesignificantlyoverthe 12-year period since 2000. the greatest contributor to the direct N2o emissions was emissions from urine and dung deposited in pasture, rangeland and paddocks, which contributed between 80.3% and 83.36% between 2000 and 2012.

Figure 5.22: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission levels of direct N2o from managed soils, 2000 – 2012


5.5.4.3 Methodological Issue

the N2o emissions from managed soils were calculated by using the tier 1 method from the ipCC 2006 Guidelines (equation 11.1). as in the 2004 agricultural inventory (DaFF, 2010), the contribution of N inputs from FSoM (N mineralization associated with loss of SoM resulting from change of land use or management) and FoS (N from managed organic soils) was assumed to be minimal and was therefore excluded from the calculations. Furthermore, sincetherearenofloodedricefieldsinSouthAfricatheseemissionswerealsoexcluded.Thesimplifiedequationfordirect N2o emissions from soils is therefore as follows:

N2oDirect-N = N2o-NN inputs + N2o-Nprp

(eq. 5.28)

Where:

N2o-NN inputs = [(FSN +FoN + FCr) * eF1] (eq. 5.29)

N2o-Nprp = [(Fprp,Cpp * eF3prp,Cpp) + (Fprp,So * eF3prp,So)]

(eq. 5.30)

Where:

N2oDirect-N = annual direct N2o-N emissions produced from managed soils (kg N2o-N yr-1);

N2o-NN inputs = annual direct N2o-N emissions from N inputs to managed soils (kg N2o-N yr-1);

N2o-Nprp = annual direct N2o-N emissions from urine and dung inputs to grazed soils (kg N2o-N yr-1);

FSN = annual amount of synthetic fertilizer N applied to soils (kg N yr-1);

FoN = annual amount of animal manure, compost, sewage sludge and other organic N additions applied to soils (kg N yr-1);

FCr = annual amount of N in crop residues, includingN-fixingcrops,andfromforage/ pasture renewal, returned to soils (kg N yr-1);

Fprp = annual amount of urine and dung N deposited by grazing animals on pasture, range and paddock (kg N yr-1), Cpp = Cattle, poultry and pigs, So = Sheep and other;

eF1 = emission factor for N2o emissions from N inputs (kg N2o-N (kg N input)-1);

eF3prp = emission factor for N2o emissions from urine and dung N deposited on pasture, range and paddock by grazing animals (kg N2o-N (kg N input)-1), Cpp = Cattle, poultry and pigs, So = Sheep and other.

Mostofthecountryspecificdatawasobtainedfromnational statistics from the Department of agriculture (abstracts of agricultural Statistics, 2012), and supporting datawasobtainedthroughscientificarticles,guidelines,reports or personal communications with experts as discussed below.

5.5.4.3.1 Nitrogen inputs

Synthetic fertilizer use (FSN) was recorded by the Fertilizer association of South africa, but organic nitrogen (FoN) and crop residue (FCr) inputs needed to be calculated. FoN is composed of N inputs from managed manure (FaM), compost and sewage sludge. FaM includes inputs from manure which is managed in the various manure management systems (i.e. lagoons, liquid/slurries, or as drylot, daily spread, or compost). the amount of animal manure N, after all losses, applied to manage soils or for feed, fuel or construction was calculated using equations 10.34 and 11.4 in the ipCC 2006 guidelines.



the amount of compost used on managed soils each year was calculated in the same way as in the 2004 inventory (DaFF, 2010). the synthetic fertilizer input changed each year, while the rest of the factors were assumed to remain unchanged over the 12 year period.

application of sewage sludge to agricultural land is common practice in South africa; however, no national data of total production of sewage sludge for South africa exists, therefore estimates were made from wastewater treatment plant data (DaFF, 2010). to estimate total sewage sludge production, a list of wastewater treatment plants (WWtp) was obtained from the Department of Water affairs (DWa, 2009) and an average capacity for each province was used to calculate the volume of wastewater treated in South africa each year. Supporting references show that 0.03% of wastewater typically could be precipitated as sewage sludge (0.1% of wastewater is solids, of which 30% is suspended) (environment Canada, 2009; van der Waal, 2008). Snyman et al. (2004) reported several end uses for sewage sludge and from this it was estimated that about 30% is for agricultural use. Due to limited data, for the years 2000 to 2004 the amount of sewage sludge used for agriculture was kept constant.

the 2004 report (DaFF, 2010), however, indicated that this amount was probably an overestimate, as the use of sewage sludge for agricultural purposes had reduced significantlyduetocontamination.Therearenofiguresonhow much this has been reduced by, so an estimated 15% reduction each year between 2004 and 2012 was assumed.

actual data for crop residue left on land was not available so the amount was calculated as described in DaFF (2010). Theprincipalbiologicalnitrogenfixing(BNF)cropsinSouth africa are soybeans, groundnuts and lucerne. the addition of N through BNF crops were included in the default values for crop-residue calculations.

5.5.4.3.2 Nitrogen inputs from urine and dung

Manure deposited in pastures, rangelands and paddocks include all the open areas where animal excretions are

not removed or managed. this fraction remains on the land, where it is returned to the soil, and also contributes to GHG emissions. in South africa the majority of animals spend most or part of their lives on pastures and rangelands. the annual amount of urine and dung N deposited on pastures, ranges or paddocks and by grazing animals (Fprp) was calculated as in the 2004 inventory using equation 11.5 in the ipCC 2006 Guidelines (Chapter 11, volume 4).


5.5.4.4.1 Synthetic fertilizer inputs (FSN)

National consumption data for N fertilizer was obtained from the Fertilizer Society of South africa (http://www.fssa.org.za).

5.5.4.4.2 Organic nitrogen inputs (FON)

Managed manure inputs (FAM)

the calculation of FaM required the following activity data:

• population data (Section 5.3.3.4.1);

• Nex data (Section 5.3.4.3.2);

• Manure management system usage data (Section 5.3.4.4.1);

• amount of managed manure nitrogen that is lost in each manure management system (FraclossMS). ipCC 2006 default values were used here (table 10.23, Chapter 10, volume 4 , ipCC 2006);

• amount of nitrogen from bedding. there were no data available for this so the value was assumed to be zero; and

• the fraction of managed manure used for feed, fuel, orconstruction.Againtherewereinsufficientdataand thus FaM was not adjusted for these fractions (ipCC 2006 Guidelines, p. 11.13).


Compost

to estimate N inputs from compost FSN data were used. it was estimated that a total of 5% of all farmers use compost. Compost is seldom, if ever, used as the only nutrient source for crops or vegetables. it is used as a supplement for synthetic fertilizers, and it is estimated that farmers would supply about a third (33%) of nutrient needs through compost (DaFF, 2010). all of this was taken into account when estimating N inputs from compost. Further details are provided in DaFF (2010) and otter (2011).

Sewage sludge

Waste water treatment data was obtained from the DWa (2009) and the end users of sewage sludge were determined from Snyman et al. (2004). Calculations follow those described in DaFF (2010) and outputs are given in otter (2011).

5.5.4.4.3 Nitrogen inputs from crop residue (FCR)

actual data for crop residue left on land was not available so the amount was calculated as described in DaFF (2010) using the following activity data:

• Crop production data from the abstract of agricultural Statistics (2012). For some of the crops the data was for a split year (i.e. July 2005/June 2006), while for others it was for a year (January to December). in order to standardize the data it was assumed that the production was evenly split between the two years and thus production per full year (January to December) was calculated. For dry peas production declined until there were no reported data between 2007 and 2012, so dry pea production was assumed to be zero for those years. the same applied to lentil production between 2006 and 2012. rye production data were not recorded between 2003 and 2009, but the data for these years werecalculatedbyfittingalinearregressiontothe

data collected since 1970.

• ipCC 2006 default values for above- and below-ground residues. in the absence of ipCC default values, external references were used to calculate the biomass by using the harvest index of crops;

• above-ground biomass removal from grazing and burning. as in the previous inventory it was assumed that all below-ground plant material remains in the soil, and that most (80%) above-ground plant material is removed by grazing. removal of above-groundbiomassincludesgrazing(forallfieldcrops)and burning (assumed only to occur with sugarcane).

5.5.4.4.4NitrogeninputsfromManuredepositedby livestock on pastures, rangelands and paddocks (FPRP)

the activity data required for this calculation were population data (Section 5.3.3.4.1), Nex data (Section 5.3.4.3.2) and the fraction of total annual N excretion for each livestock species that is deposited on pastures, rangelands and paddocks (prp).


the ipCC 2006 default emission factors (Chapter 11, volume 4, table 11.1) were used to estimate direct N2o emissions from managed soils. eF1 was used to estimate direct N2o-N emissions from FSN, FoN and FCr N inputs; while eF3prp was used to estimate direct N2o-N emissions from urine and dung N inputs to soil from cattle, poultry and pigs (Cpp), and sheep and other animals (So).

For game the default factor for other animals (i.e. the So eF) was used.

the ipCC 2006 default eFs for prp were thought to be overestimated for South africa, as grazing areas in South africa are mostly in the drier parts of the country where water content is low. even though the N is available as a potential source of N2o, this is not the most likely



pathway. the 2004 inventory (DaFF, 2010) suggests that emissions from prp are probably more towards the lower range of the default values provided by the ipCC (2006).



the uncertainty ranges for eF1, eF3prp,Cpp and eF3prp,So are 0.003 to 0.03, 0.007 to 0.06, and 0.003 to 0.03, respectively (ipCC 2006 Guidelines, table 11.1).

the uncertainty on the amount of compost applied was high due to the high variability across the region. No quantitative data was available, so estimates were based on expert opinion. Most estimates indicated a use of 20% animal manure in producing compost, while some indicated as much as 80%. Most vegetable production is estimated to use compost for about 33% of the nutrient requirements. the data could be improved if more information on organic matter use in agriculture were available.

Sewage sludge use in agriculture is also uncertain due to a lack of actual data. in this inventory, data from 72 WWtps were used as representative of all 1 697 WWtps in South africa. a calculated value based on estimated capacity ranges for wastewater treatment was used to determine the total quantity of sewage sludge produced. ranges of WWtp capacity varied between 1 776 Ml and 8 580 Ml per day, with an average of 5 177 Ml per plant. the end use of sewage sludge could also be improved. it was estimated that 30% of total sewage sludge is used in the agricultural sector, but this number could vary between 10 and 80%. the N content for sludge, from the 72 WWtps data, varied between 1.5% and 6.5%. the average of 3.8% was used in the calculations.

the main source of data for crop residue was the abstract of agricultural Statistics (2012), and one of the main limitations was that it only included commercial crops, not subsistence agriculture. For the rest of the calculation,

the default values were mostly used, or estimates of the fraction of crop removal. these estimates were often area-specificbutappliedonabroadscale.Infuture,eachprovince or area could be documented separately to give more accurate data input.


the same data sources and emission factors were used for the 12 year period to maintain the time-series consistency over the inventory period.


Itwasdifficulttoverifyalotofthedatainthissectiondue to there being very limited data available to cross-check the numbers against. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and identify areas of improvement.


the livestock population data were updated for other cattle and poultry, therefore recalculations were done for 2000 to 2012 due to the associated change in manure production. an estimate for game manure emissions was included (which was not in previous inventories) and this amount was included in the emissions for each year since 2000.


No source-specific improvements are planned, but suggestions are given as to how estimates could be improved in future. the major component of N2o emissions from managed soils was from the inputs of animal manure and manure deposited on pastures, rangelands and paddocks. it is, therefore, very important forSouthAfricatodeterminecountry-specificEFsfor


these emissions, instead of using the ipCC default values. Furthermore, as with the CH4 emissions from manure management, the manure-management system usage dataneedstobebetterquantifiedinordertoimprovethe accuracy of the emission estimates in this category. lastly it is recommended that FSoM (N mineralization associated with loss of SoM resulting from change of land use or management) and FoS (N from managed organic soils) be investigated further. these sources should be regarded as areas for potential improvement and attempts should be made to include these emissions in future.

5.5.5 Indirect N2O emissions from managed soils [3C4]


indirect emissions of N2o-N can take place in two ways:

i) volatilization of N as NH3 and oxides of N, and the deposition of these gases onto water surfaces, and ii) through runoff and leaching from land where N was applied (ipCC, 2006). Due to limited data a tier 1 approach was used to calculate the indirect N2o emissions.


the total accumulated amount of indirect N2o lost over the period 2000 to 2012 was estimated at 60 358 Gg Co2eq and 4 558 Gg Co2eq was produced in 2012. there was a decreasing trend over this period with losses due to atmospheric deposition of N volatilised from managed soils decreasing by 7.5% and losses due to leaching and run-off declining by 6.0% between 2000 and 2012 (Figure 5.23). indirect N2o losses due to leaching and run-off accounted for 56.5% of the indirect emissions in 2012.

Figure 5.23: Sector 3 aFolU – aggregated and non-Co2 sources: trend and emission level estimates of indirect N2o losses from managed soils, 2000 – 2012.




a tier 1 approach was used to estimate the indirect N2o losses from managed soils. the annual amount of N2o-N produced from atmospheric deposition of N volatilized from managed soils (N2o(atD)-N) was calculated using ipCC 2006 equation 11.9; while equation 11.10 was used to estimate the annual amount of N2o-N produced from leaching and runoff of N additions to managed soils (N2o(l)-N) (Chapter 11, volume 4, ipCC 2006).


the values for FSN, FoN, Fprp, FCr, and FSoM were taken from section 5.5.4.4 of this report. the emission (eF4 and eF5), volatilization (FracGaSF and FracGaSM) and leaching (FracleaCH-(H)) factors were all taken from the ipCC 2006 default table (table 11.3, Chapter 11, volume 4, ipCC 2006).



there is uncertainty in the activity data; nevertheless emission factor uncertainty is likely to dominate. the uncertainty ranges on eF4 and eF5 are 0.002 to 0.05, and 0.0005 to 0.025, respectively. For FracGaSF, FracGaSM and FracleaCH-(H), the uncertainty ranges are 0.03 to 0.3, 0.05 to 0.5 and 0.1 to 0.8, respectively (ipCC 2006 Guidelines, table 11.3, p. 11.24).


the same data sources were used throughout the 12 year period to reduce uncertainties due to inconsistent data sources.


Nosource-specificQA/QCandverificationprocedures

were carried out in this section.


recalculations were completed for all years between 2000 and 2010 due to adjustments in the livestock population numbers (other cattle, poultry and game) which led to changes in the Fprp inputs.


Nospecificimprovementshavebeenplannedforthissource.

5.5.6 Indirect N2O emissions from manure management [3C6]


indirect N2o losses from manure management due to volatilization were calculated using the tier 1 method. throughout the world, data on leaching and run-off losses from various management systems is extremely limited, therefore, there are no ipCC 2006 default values and no tier 1 method. the equation given in the ipCC 2006 Guidelines can only be used where there is country-specific information on the fraction of nitrogen loss due to leaching and run-off from manure management systems available, i.e. there is only a tier 2 method. TherewasinsufficientdataforSouthAfricatodothetier 2 calculation, so there is no estimate for manure management N losses due to leaching and run-off.


the amount of manure N lost due to volatilized NH3 and Nox was calculated as described in the ipCC 2006 Guideline default equations and emission factors. the total accumulated loss in N2o from manure was estimated


at 5 081 Gg Co2eq between 2000 and 2012. the annual variation was low and there was an increasing trend from 349 Gg Co2eq in 2000 to 436 Gg Co2eq in 2012 (Table 5. 20).


the ipCC 2006 Guideline tier 1 approach was used to estimate N2o losses due to volatilization from manure management.


the amount of manure N lost due to volatilized NH3 and Nox was calculated using equation 10.26 (ipCC, 2006). this requires Nex data (Section 5.3.3.4.1), manure management system data (Section 5.3.4.4.1), and default fractions of N losses from manure management systems due to volatilization ((ipCC 2006, table 10.22) (Table 5. 21).


a default emission factors for N2o from atmospheric deposition of N on soils and water surfaces (given in the ipCC 2006 Guidelines as 0.01 kg N2o-N (kg NH3-N + Nox-N volatilized)-1) was used to calculate indirect N2o emissions due to volatilization of N from manure management (equation 10.27, ipCC 2006).



the uncertainty on N losses from manure management

systems due to volatilization was high because of the wide ranges on default values (see Table 5. 21) and uncertainty on manure management system usage. the uncertainty range on eF4 is 0.002 – 0.05 (ipCC 2006 Guidelines, table 11.3).


the same data sources and emission factors were used throughout the 12 year period to ensure time-series consistency.


there was no previous data to compare the values with, makingqualitycontrolverydifficult,sonosource-specificQa/QC was done on this category. an independent reviewer was appointed to assess the quality of the inventory, determine the conformity of the procedures followed for the compilation of this inventory and to identify areas of improvements.


Nosource-specificrecalculationswerenecessary.


the indirect N2o emissions from manure management form a very small component of the overall N2o emission budget and so there are no immediate plans to improve this section.

table 5.20: Sector 3 aFolU – aggregated and non-Co2 sources: indirect emissions of N2o (Gg Co2eq) due to volatilization from manure management between 2000 and 2012.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

indirect N2o

349 352 358 360 364 382 397 411 410 412 414 428 436



5.5.7 Harvested wood products


Much of the wood that is harvested from forest land, cropland and other land types remains in products for differing lengths of time. this section of the report estimates the contribution of these harvested wood products (HWps) to annual Co2 emissions or removals. HWps include all wood material that leaves harvest sites.


HWps were estimated to be a sink of Co2 (except in 2011 whenitwasaweaksource)whichfluctuatedannuallybetween 2000 and 2012. the total accumulated sink over the 12 years was estimated at 7 376 Gg Co2 with a sink

of 512 Gg Co2 in 2012. the sink increased from 312 Gg Co2 in 2000 to 1 185 Gg Co2 in 2004. after this the sink fluctuatedbetween98GgCO2 and 882 Gg Co2 with a source of 82 Gg Co2 in 2011 (Table 5. 22).


the HWp contribution was determined by following the updated guidance provided in the 2013 ipCC Kp Supplement (ipCC, 2014). one of the implications of Decision2/CMP.7isthataccountingofHWPisconfinedto products in use where the wood was derived from domestic harvest. Carbon in imported HWp is excluded. the guidelines also suggest that it is good practice to allocate the carbon in HWp to the activities afforestation (a), reforestation (r) and deforestation (D) under article 3 paragraph 3 and forest management (FM) under article 3paragraph4.ForSouthAfrica,thereisinsufficientdata

table 5.21: Sector 3 aFolU – aggregated and non-Co2 sources: Default values used for N loss due to volatilization of NH3 and Nox from manure management (%). the value in the brackets indicates the range.

Livestock Category

Lagoon Liquid /slurry Drylot Daily spread Compost

Dairy Cattle 35 (20-80) 40 (15-45) 20 (10-35) 30 (10-40)

Commercial Beef Cattle

30 (20-50) 45 (10-65)

Subsistence Cattle

30 (20-50) 45 (10-65)

Sheep 25 (10-50)

Goats 25 (10-50)

Horses

Donkeys

pigs 40 (25–75) 48 (15-60) 25 (10-50) 45 (10-65) 25 (15-30)

poultry 55 (40-70) 55 (40-70)

table 5.22: Sector 3 aFolU – other: trend in the HWp Co2 sink (Gg) between 2000 and 2012.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

HWp -312 -675 -817 -927 -1 185 -197 -882 -581 -781 -98 -490 82 -512

Columns that have no data are not required as there is no manure management in this division for this livestock.


to differentiate between the harvest from ar and FM, it is conservative and in line with good practice to assume that all HWps entering the accounting framework originate from FM (Kp Supplement, Chapter 2, p 2.118).

equation 5.31 and 5.32 (eq 2.8.1 and 2.8.2 in Kp Supplement) were applied to estimate the annual fraction of feedstock for HWp production originating from domestic harvest and domestically produced wood pulp as feedstock for paper and paperboard production.

firW(i) = (irWp(i) – irWex(i))/ (irWp(i) + irWiM(i) – irWex(i))

(eq. 5.31)

Where:

firW(i) = share of industrial roundwood for the domestic production og HWp originating from domestic forests in year i;

irWp(i) = production of industrial roundwood in year i (Gg C yr-1);

irWiM(i) = import of industrial roundwood in year i (Gg C yr-1);

irWex(i) = export of industrial roundwood in year i (Gg C yr-1).

fpUlp(i) = (pUlpp(i) – pUlpex(i))/ (pUlpp(i) + pUlpiM(i) – pUlpex(i))

(eq. 5.32)

where:

fpUlp(i) = share of domestically produced pulp for the domestic production of paper and paperboard in year i;

pUlpp(i) = production of wood pulp in year i (Gg C yr-1);

pUlpiM(i) = import of wood pulp in year i (Gg C yr-1);

pUlpex(i) = export of wood pulp in year i (Gg C yr-1).

the resulting feedstock factors were applied to equation 5.33 (eq 2.8.4 Kp Supplement) to estimate the HWp contribution of the aggregate commodities sawnwood, wood-based panels and paper and paperboard.

HWpj(i) = HWpp(i) * fDp(i) * fj(i)

(eq. 5.33)

where:

HWpj(i) = HWp amounts produced from domestic harvest associated with activity j in year i (m3 yr-1 or Mt yr-1);

HWpp(i) = production of the particular HWp commodities (i.e. sawnwood, wood-based panels and paper and paperboard) in year i (m3 yr-1 or Mt yr-1);

fDp(i) = share of domestic feedstock for the production of the particular HWp category originating from domestic forests in year i, with:

fDp(i) = f irW(i) for HWp categories ‘sawnwood’ and ‘wood-based panels’; and

fDp(i) = firW(i) * fpUlp(i) for HWp category ‘paper and paperboard’; and

firW(i) = 0 if firW(i) < 0 and fpUlp(i) = 0 if fpUlp(i) < 0.

fj(i) = share of harvest originating from the particular activity j (FM or ar or D) in year i. For Sa this was assumed to be 1 as all the harvest was allocated to FM.

First order decay

transparent and verif iable data were available for sawnwood, wood-based panels and paper and paperboard, butnocountry-specificinformationforTier3wasavailablesoaTier2firstorderdecayapproach(Eq5.34(Eq12.1



in 2006 ipCC Guidelines)) was applied to estimate the HWp contribution:

C(i+1) = e-k * C(i) + ((1 – e-k)/k) * Inflow(i)

(eq. 5.34)

Where:

C(i) = the carbon stock in the particular HWp category at the beginning of year i (Gg C);

k = decay constant of FoD for each HWp category (units yr-1) (k = ln(2)/Hl where Hl is the half life of the HWp pool in years;

Inflow(i) =theinflowtotheparticularHWP category during year i (Gg C yr-1);

ΔC(i) = C(i+1) – C(i) = carbon stock change of the HWp category during year i (Gg C yr-1).

as a proxy in the tier 2 method it is assumed that the HWp pools are in steady state at the initial time (t0) from which the activity data start. this means that as a proxy ΔC(t0)isassumedtobeequalto0andthissteadystatefor each HWp commodity category is approximated using the following equation (eq 2.8.6 Kp Supplement):

C(t0) = Inflowaverage/k

(eq. 5.35)

where:

Inflowaverage=(∑t4 Inflow(i))/5

(eq. 5.36)

C(t0) was taken to be 1990 (S. ruter, pers. comm.) and was substituted into eq 5.34 so that C(i)andΔC(i) in the sequential time instants can be calculated.


all the data on production, imports and exports of roundwood, sawnwood, wood-based panels, paper and paperboard, and wood pulp were obtained from the FaoStat database (http://faostat.fao.org/).


the activity data was obtained from the Fao and the same data set, dating back to 1961, was applied throughout to maintain consistency. Uncertainties for activity data and parameters associated with HWp variables are provided in the ipCC Guidelines (ipCC 2006, volume 4, p. 12.22). production and trade data have an uncertainty of 50% since 1961, while the product volume to product weight factors and oven-dry product weight to carbon weight have uncertainties of ±25% and ±10%, respectively. there was also a ±50% uncertainty on the half-life values.


as part of the quality control the data was run through the WoodCarbonMonitor model and the ipCC HWp model and the outputs were compared. although there were some slight differences the data were all within a similar range.


recalculations were necessary as the recent Kp Supplement guidelines were applied, so as to reduce uncertainty, and there were several changes to the methodology. Data was recalculated for all years between 2000and2012.Recalculationsresultedinasignificantreduction in the Co2 sink of the HWp category.


Due to the recent improvements made to this subsector therearenofurthersource-specificimprovementsplanned.

i=t0


6.1 Overview of Sector

Climate change caused by greenhouse gas (GHG) emissions, mainly from anthropogenic sources, is one of themostsignificantchallengesdefininghumanhistoryover the past few decades. among the sectors that contribute to the increasing quantities of GHGs into the atmosphere is the waste sector. this section highlights the GHG emissions into the atmosphere from managed landfillsandwastewater treatmentsystems inSouthafrica, estimated using the ipCC 2006 Guidelines.

the waste sector in the national inventory of South africa comprises two sources:

• 4a Solid waste disposal; and

• 4D Wastewater treatment and discharge.

the results were derived by either using available data or estimates based on accessible surrogate data sourced fromthescientificliterature.Forthewastesector,amongthe chief limitations of quantifying the GHG emissions from different waste streams was the lack of a periodically updated national inventory on: the quantities of organic wastedepositedinwell-managedlandfills;theannualrecoveryofmethanefromlandfills;quantitiesgeneratedfrom anaerobically decomposed organic matter from wastewater treated; and per capita annual protein consumption in South africa.

Tocontextualizethefindingspresentedhereinandprovidea sound basis for interpreting them, the assumptions used in estimating the 2000 – 2010 GHG emissions from the waste sector (Department of environmental affairs and tourism, 2014) were adopted for this 2000 – 2012 inventory. therefore, the entire set of assumptions will not be reproduced in this report. However, even though a large percentage of the GHG emissions from waste sources are expected to come from managed solid wastelandfillsandwastewatertreatmentsystems,future

inventories should address completeness in this sector by quantifying emissions from the following sources:

• the open burning of waste, as these emissions also have potential impacts for air-quality management;

• the biological treatment of organic waste, where a clear and unambiguous link with agricultural practices can be made; and

• the incineration of solid waste and biological waste.

6.2 Overview of shares and trends in emissions

the total estimated GHG emissions from the waste sector were estimated to have increased by 78.5% from 12 288 Gg Co2eq in 2000 to 21 928 Gg Co2eq in 2012 (Figure 6.1). the annual increase declined from 7.5% to 3.7% between these years. emissions from solid waste disposal dominated (Figure 6.1) with its contribution to the total GHG emissions from the waste sector increasing steadily from 76.2% in 2000 to 84.0% in 2012. there aretwolikelyreasonsfortheincrease:firstly,thefirstorder decay (FoD) methodology has an in-built lag-effect and, as a result, the reported emissions from solid waste inmanagedlandfillsinagivenyeararelikelytobedueto solid waste disposed of over the previous 10 to 15 years. Secondly, in South africa the expected growth in the provision of sanitation services, particularly with respect to collecting and managing solid waste streams inmanaged landfills, is likely toresult inan increasein emissions of more than 5% annually. in addition, at present very little methane is captured at the country’s landfillsandthepercentagesofrecycledorganicwastearelow. intervention mechanisms designed to reduce GHG emissionsfromsolidwastearelikelytoyieldsignificantreductions in the waste sector.

6. WASTE SECTOR

6. Waste Sector


6.3 Key categories

the key categories in the waste sector are shown to be wastewater treatment and discharge (domestic) (4.D.1) in terms of level (Table 1.9), and in terms of the trend solid waste disposal (4.a) is a key (Table 1.11).

6.4 Solid waste disposal on land [4A]

6.4.1 Source-category description

in 2000 it was estimated that the disposal of solid waste contributed less than 2% of the total GHG emissions in South africa, mainly through emissions of methane from urbanlandfills(DepartmentofEnvironmentalAffairsandtourism 2009). Waste streams deposited into managed landfillsinSouthAfricacomprisewastefromhouseholds,

Figure 6.1: Sector 4 Waste: trends and emission levels of source categories, 2000 – 2012.


commercial businesses, institutions, and industry. in this report only the organic fraction of the waste in solid disposal sites was considered as other waste stream components were assumed to generate insignificant quantitiesinlandfills.FurthermoreonlyGHG’sgeneratedfrommanageddisposal landfills inSouthAfricawereincluded, as data on unmanaged sites are not documented and the sites are generally shallow. a periodic survey is still needed to assess the percentage share of unmanaged sites and semi-managed sites. Generating this information is central to understanding methane generation rates for different solid waste disposal pathways.


the total accumulated GHG emissions from solid waste disposal between 2000 and 2010 was estimated at 183 048 Gg Co2eq, increasing from 9 368 Gg Co2eq in 2000 to 18 412 Gg Co2eq in 2012 (Figure 6.1). this is an increase of 78.5% over the 12 year period.


the FoD model was used to estimate GHG emissions from this source category for the period 1950 to 2012 using the following input activity data: population, waste generation rates, income per capita, annual waste generation and population growth rates, emission rates, half-lives of bulk waste stream (default value for the half-live is 14 years), rate constants, methane correction factor (MCF), degradable carbon fraction (DCF), as well as other factors described in the ippC Guidelines, volume 5, Chapter (ippC, 2006). Notably, due to a lack of published specific-activitydataformanyoftheseparametersinSouth africa, the default values suggested in the ipCC Guidelines were applied. in addition, the inventory compilers noted that the information on national waste composition presented in the National Waste Baseline information report (NWiBr, 2012) was not compatible with the approach set out in the 2006 ipCC Guidelines,

therefore, even though domestic information on waste composition was available, it could not be used for the purposes of this inventory. instead, default ipCC waste composition values were used. the NWiBr, however, does provide information on the percentage of the waste generated that is disposed of in solid waste disposal sites. this value was estimated at 91% for 2011. it was assumed that this value was constant throughout the time series due to a lack of data for the previous years.

6.4.4 Data sources

For the FoD methodology, the model required historical datawithatleastthreetofivehalf-lives.Therefore,theactivity data used comprised waste quantities disposed ofintomanagedlandfillsfrom1950to2010,coveringa period of about 70 years (satisfying the condition foraperiodoffivehalf-lives).Populationdatafortheperiod 1950- to 2001 was sourced from United Nations population statistics (UN, 2012). Statistics South africa population data was used for the period 2002 to 2012 (StatsSa, 2013). Waste Generation rates for industrial waste were estimated using GDp values sourced from StatsSa (StatsSa, 2013) and tonnages of industrial waste reported in the NWiBr for 2011. a value of 8 tonnes/per unit of GDp in US dollars was derived and assumed constant throughout the time series.

among the chief limitations of the FoD methodology is that even if activity data improved considerably, the limitations of the data, or lack thereof, of previous years will still introduce a considerable degree of uncertainty. on the other hand, the estimated waste generations derived from previous years, back to 1950, will remain useful in future estimations of GHGs as they will aid in taking into account half-lives.

No detailed analysis of the methane recovery from landfills was accounted for between 2000 and 2010. as noted in the previous inventory (Department of environmental affairs, 2009), the recovery of methane fromlandfillscommencedonalarge-scaleafter2000,

6. Waste Sector


with some sites having a lifespan of about 21 years (DMe, 2008). to address these data limitations, the Dea has implemented the National Climate Change response Database, which captures valuable data from mitigation and adaptation projects for future GHG estimates from landfills.Thistoolwillbeusedinthefuturetoidentifyand implement methane recovery projects. However, at present there are limited publicly accessible data on the quantities of methane recovered annually from managed landfillsinSouthAfrica.


6.4.5.1 Uncertainty

Uncertainty in this category is due mainly to the lack ofdataonthecharacterizationof landfills,aswellasof the quantities of waste disposed in them over the medium to long term. an uncertainty of 30% is typical for countries that collect waste-generation data on a regular basis (ipCC 2006 Guidelines, table 3.5). another source of uncertainty is that methane production is calculated using bulk waste because of a lack of data on waste composition, therefore, uncertainty is more than a factor of two (Department of environmental affairs and tourism, 2009). For the purpose of the bulk waste estimates,thewholeofSouthAfricaisclassifiedasa“warm dry temperate” climate zone, even though some landfillsarelocatedindrytropicalclimaticconditions.other uncertainties are the fraction of MSWt sent to SWDS (more than a factor of two), DoCf (±20%), MCF (-10%-0%), F (±5%), methane recovery (can be as much as ±50%) and the oxidation factor (ipCC 2006 Guidelines, table 3.5).


the FoD methodology for estimating methane emissions from solid waste requires a minimum of 48 years’ worth of historical waste disposal data. However, in South africa,

waste disposal statistics are not available. in addition, periodic waste baseline studies do not build time-series data. Hence, population statistics sourced from the UN secretariat provided consistent time-series activity data for solid waste disposal.


a review of the waste sector emission estimates has been performed by experts from various universities. the review resulted in major changes to emission estimates. For example, assumptions about the percentage of waste that ends in waste disposal sites, GDp values for estimating emissions from industrial waste, as well as waste generation rates were all reviewed and updated. Verificationfocusedonwastegenerationrates,populationstatistics and GDp values using Statistics Sa datasets and the Dea’s waste baseline study.



Source-specificrecalculationshavebeenperformedforthe period 2000 to 2010 in this source category because of the following:

• population statistics for the period 2002 to 2010 based on new information available in the 2011 national census conducted by Statistics South africa.

• the percentage of waste generated that is disposed in solid waste disposal sites was reviewed and updated based on the NWiBr.

• a waste generation rate for municipal solid waste was reviewed based on new information sourced from the NWiBr. the new waste generation rate is based on provincial weighted average waste generation rates. this rate was assumed constant for the entire time period.


• a waste generation rate for industrial waste was reviewed based on GDp data reported by StatsSa as well as industrial waste tonnage rates available from the NWiBr for 2011. this rate was assumed constant for the entire time period.


the most challenging task in estimating GHG emissions in South africa was the lack of specific-activity and emissions factor data. as a result, estimations of GHG emissions from both solid waste and wastewater sources were largely computed using default values suggested in ipCC 2006 Guidelines and, as a consequence, margins of error were large. therefore, several recommendations for improving the activity and emission factors data are suggested. these include:

(i) advance data capturing, particularly the quantities of wastedisposedofintomanagedandunmanagedlandfills.other activities that merit improvement are the MCF and rate constants, owing to their impact on the computed methaneemissionsfromlandfills;and

(ii) improve the reporting of economic data (e.g. annual growth) to include different population groups. the assumption that GDp growth is evenly distributed (using a computed mean) across all the population groups is highly misleading and leads to exacerbated margins of error.

6.5 Wastewater treatment and discharge [4D]


Wastewater treatment contributes to anthropogenic emissions, mainly CH4 and N2o. the generation of CH4 is due to anaerobic degradation of organic matter in

wastewater from domestic, commercial and industrial sources. the organic matter can be quantified using biological oxygen demand (BoD) values.

Wastewater can be treated on site (mostly industrial sources), or treated in septic systems and centralised systems (mostly for urban domestic sources), or disposed of untreated (mostly in rural and peri-urban settlements). Most domestic wastewater CH4 emissions are generated from centralised aerobic systems that are not well managed, or from anaerobic systems (anaerobic lagoons and facultative lagoons), or from anaerobic digesters where the captured biogas is not completely combusted.

Unlike in the case of solid waste, organic carbon in wastewater sources generates comparatively low quantities of CH4. this is because even at very low concentrations, oxygen considerably inhibits the functioning of the anaerobic bacteria responsible for the generation of CH4.

N2Oisproducedfromnitrificationanddenitrificationof sewage nitrogen, which results from human protein consumption and discharge.

the revised 1996 ipCC Guidelines (ipCC 1997) included one equation to estimate emissions from wastewater and another to estimate emissions from sludge removed from wastewater. this distinction was removed in the 2006 ipCC Guidelines (ipCC 2006, vol.5, p 6.9), so both emissions are now calculated by the same equation.

in South africa, most of the wastewater generated from domestic and commercial sources is treated through municipal wastewater treatment systems (MWtps).

For wastewater generated by industrial processes, the ipCC 2006 Guidelines list the industry categories which use large quantities of organic carbon that generate wastewater (ipCC 2006 vol.5, p.6.22). the ipCC 2006 Guidelines require the development of consistent data for estimating emissions from wastewater in a given industrial sector (ipCC 2006, vol.5, p 6.22). once an industrial

6. Waste Sector


sector is included in an inventory, it should be included in all future inventories.

the South african data on industrial categories with high organic content are very limited. Some data exist on wastewater in sectors such as vegetables, fruits and juices, and the wine industry, but these are available only foraspecificyear,makingitimpossibletoextrapolatesuch statistics accurately over any period. therefore, in this inventory, only CH4 emissions from domestic sources are presented. However, wastewater from commercial and industrial sources discharged into sewers is accounted for, so the term “domestic wastewater” in this inventory refers to the total wastewater discharged into sewers from all sources. this is achieved by employing the default ipCC methane correction factor (MCF) of 1.25 used to account for commercial and industrial wastewater. it is highly likely that the MCF value for South africa ranges between 1.2 and 1.4.

Domestic and commercial wastewater CH4 emissions mainly originate from septic systems and centralised treatment systems such as MWtps. Because of the lack of national statistics on the quantities of BoD generated from domestic and commercial sources in South africa annually, the yearly estimates were determined using the ippC 2006 default tier 1 method.


Domestic and commercial waste water treatment and discharge were estimated to produce a total accumulated emission of 41 362 Gg Co2eq between 2000 and 2012. the CH4 emissions accounted for approximately 80.4% of total emissions (Table 6.1). in 2012 CH4 emissions totalled 2827 Gg Co2eq, while N2o emissions contributed 689 Gg Co2eq (Table 6.1).

table 6.1: Sector 4 Waste: CH4 and N2o emissions from domestic and industrial wastewater treatment, 2000 – 2012.

Wastewater Treatment and Discharge CH4 emissions

Wastewater Treatment and Discharge N2O emissions

Total GHG emission

(Gg Co2eq)

2000 2 348 572 2 921

2001 2 397 584 2 981

2002 2 430 592 3 022

2003 2 461 600 3 061

2004 2 491 607 3 097

2005 2 519 613 3 132

2006 2 548 621 3 168

2007 2 576 628 3 203

2008 2 604 635 3 239

2009 2 632 641 3 274

2010 2 661 649 3 310

2011 2 763 673 3 436

2012 2 828 689 3 516


the urban low-income population had the highest total contribution of methane emissions. results suggest that for South africa to reduce the methane emissions from wastewater sources, directed interventions such as increasing the low-income urban population served by closed-sewer treatment systems is critical. this is because closed sewer treatment systems are suitable for potential capturing of generated methane emissions – unlike the open latrines and sewer systems currently serving approximately 60% of the low-income urban population in South africa.


6.5.3.1 Domestic Wastewater Treatment and Discharge

the projected methane emissions from the wastewater follow the same methodology described in the 2014 National GHG inventory report (Department of environmental affairs and tourism, 2014). the estimated methane emissions reported are from domestic and commercial sources of wastewater because the ippC 2006 Guidelines do not stipulate a different set of equations or differentiated computational approaches for the two sources, as was previously stipulated in 1996 ipCC Guidelines. it should be noted that the data on quantitiesofwastewaterfromspecificindustrialsourceswith high organic content are largely lacking in South africa and, therefore, the estimated values in this report are assumed to be due to domestic and industrial sources treated in municipal wastewater treatment systems.

6.5.4 Data sources

Tobeconsistent,thespecific-categorydatadescribedinSection 6.4.1 of the National GHG inventory report (Department of environmental affairs and tourism, 2009) and its underlying assumptions were adopted. For example, in determining the total quantity of kg BoD yr-1, population data was sourced from Statistics South

africa’s mid-year population statistics published in 2013 (StatsSa, 2013). Default population distribution trends between rural and urban settlements as a function of income, as well as a default average South african BoD production value of 37 g person-1 day-1 were sourced from the 2006 ipCC Guidelines. Generally, it is good practice to express BoD product as a function of income, however, this information is not readily available in South africa, therefore, it could not be included in the waste sector model. in this case, a default ipCC correction factor of 1.25 was applied in order to take into account the industrial wastewater treated in sewage treatment systems.

the emissions factors for different wastewater treatment and discharge systems were taken from the 2006 ipCC Guidelines (Table 6.2) as was the data on distribution and utilization of different treatment and discharge systems (Table 6.3).

6.5.4.1 Domestic Wastewater N2O Emissions

the default values provided by the ipCC Guidelines were used in estimating the potential growing trends of N2o emissions from the wastewater treatment systems. this wasduetothelackofspecific-activitydataforSouthafrica. For instance, a default value for per capita protein consumption of 27.96 kg yr-1 was applied in the model.

6.5.5 Uncertainties and time-series consistency

6.5.5.1 Uncertainties

an analysis of the results for methane emissions suggest that the likely sources of uncertainties may be due to the input data. these include uncertainties associated with South african population estimates provided by the United Nations, the presumed constant country BoD production of about 37 g person-1 day-1 from 2001 to 2020, and the lack of data on the distribution of wastewater treatment systems in South africa. it is recommended

6. Waste Sector


table 6.2: Sector 4 Waste: emission factors for different wastewater treatment and discharge systems (Source: Department of environmental affairs and toursim, 2009).

Type of treatment or discharge

Maximum CH4 producing capacity

(bOD)

CH4 correction factor for each treatment

systemEmission factor

(kg CH4/kg BoD) (MCF) (kg CH4/kg BoD)

Septic system 0.6 0.5 0.30

latrine – rural 0.6 0.1 0.06

latrine - urban low income 0.6 0.5 0.30

Stagnant sewer (open and warm) 0.6 0.5 0.30

Flowing sewer 0.6 0.0 0.00

other 0.6 0.1 0.06

None 0.6 0.0 0.00

table 6.3: Sector 4 Waste: Distribution and utilization of different treatment and discharge systems (Source: Department of environmental affairs and tourism, 2009).

Income group Type of treatment or discharge pathwayDegree of utilization

(tij)

rural

Septic tank 0.10

latrine – rural 0.28

Sewer stagnant 0.10

other 0.04

None 0.48

Urban high-income

Sewer closed 0.70

Septic tank 0.15

other 0.15

Urban low-income

latrine - urban low income 0.24

Septic tank 0.17

Sewer (open and warm) 0.34

Sewer(flowing) 0.20

other 0.05


that, in future inventories, a detailed study on the input parameters merits careful consideration to minimize the uncertainty level. in turn, this approach would improve the reliability of the projected methane estimates from wastewater sources.


time-series consistency was achieved by using population datasets obtained from the UN secretariat. assumptions about wastewater streams were assumed to be constant over the 12-year time series and default ipCC emission factors used.


internal and external reviews of this source category were included in the review of solid waste disposal. Hence, changes to population statistics, the percentage split of wastewater pathways, total organics in wastewater and methane correction factors were all reviewed.


one correction was made to the calculations since the 2000 inventory. in the previous inventory individual toW (total organics in wastewater) values for each income group were used to calculate a CH4 emission from each income group which were then summed together to obtain the total emissions. Whereas in this inventory the total toW (of all the income groups) was used in the emission calculation. this correction was made following equation 6.1 in the ipCC 2006 Guidelines.


the biggest challenge in estimating GHG emissions in SouthAfricaisthelackofspecific-activityandemissionsfactor data. as a result, estimations of GHGs from both solid waste and wastewater sources were largely computed using default values suggested in the ipCC 2006 Guidelines and, as a consequence, margins of error were high. therefore, several recommendations are suggested to improve the activity and emission factors data:

• obtain data on the quantities of waste disposed of intomanagedandunmanagedlandfills;

• improve the MCF and rate constants;

• improve the reporting of economic data (e.g. annual growth) to include different population groups. the assumption that GDp growth is evenly distributed (using a computed mean) across all the population groups is highly misleading, and leads to exacerbated margins of error;

• obtain information on population distribution trends between rural and urban settlements as a function of income;

• Conduct a study to trace waste streams and obtain more information on the bucket system which is still widely used in South africa.

6. Waste Sector


7. REFERENCES

abstracts of agricultural Statistics, 2012. Department of agriculture, Forestry and Fisheries, pretoria, rSa. Can be downloaded from www.daff.gov.za.

airports Company South africa (aCSa), 2013. passenger travel Statistics. http://www.acsa.co.za/home.asp?pid=137. [accessed april 2013]

Alembong,OA.,2015.CarbonflowanalysisintheSouthafrican Forest and the Forestry Sector, MSC thesis, University of KwaZulu-Natal.

anaya, J.a. and Chuvieco, e., 2012. accuracy assessment of Burned area products in the orinoco Basin, Photogrammetric Engineering & Remote Sensing, 78: 53 – 60.

anglo american 2011: the South african iron and Steel value Chain. http://www.angloamericankumba.com/ [accessed July 2012]

anton eberhard, 2009. South African Coal: market, investment and policy challenges. South african coal paper, University of Cape town.

archibald, S., roy, D.p., van Wilgen, B.W, Scholes, r.J., 2010.Whatlimitsfire?Anexaminationofthedriversofburnt area in southern africa. GlobalChangeBiology.

archibald, S., Scholes, r.J., roy, D., roberts, G. and Boscheti, l., 2009. Southern african fire regimes as revealed by remote sensing. african Journal of ecology.

association of Cementitious Material producers, 2008. Sustainability report. http://www.acmp.co.za/climate_change.htm [accessed March 2013

association of Cementitious Material producers, 2009. Cement Manufacturing Process. http://www.acmp.co.za/cement_manufacturing_process.htm. [accessed March 2013].

aucamp, a.J. and Howe, l.G., 1979. the production potential of the eastern Cape valley Bushveld. Proc. Grassl. Soc. S Afr. 14: 115 – 117.

australian National inventory report (aNir), 2009. australian national greenhouse accounts: National inventory report. Department of climate change and energyefficiency,CommonwealthofAustralia,Canberra,aCt.

Baggott, S.l., Brown, l., Cardenas, l., Downes, M.K., Garnett, e., Hobson, M., Jackson, J., Milne, r., Mobbs, D.C., passant, N., thistlethwaite, G., thomson, a. and Watterson, J., 2006. Greenhouse Gas inventory, 1990 to 2004.

Bessinger, D., 2000. Cooling Characteristics of High Titania Slags, MSc dissertation, University of pretoria, pretoria. http://upetd.up.ac.za/thesis/available/etd-07212006-102324 [accessed 30 January 2013].

Bexley, J., 2013. A perspective on the Glass Manufacturing Industry. (personal communication, February, 2013).

Blaxter, K.l. and Clapperton, J.l., 1965. prediction of the amount of methane produced by ruminants. Brit. J. Nutr. 19: 511 – 522.

Boschetti, l., roy, D., Hoffman, a.a. and Humber, M., 2012. MoDiS Collection 5.1 burned area product – MCD45: User’s guide version 3.

Brouwer, 1965. report of sub-committee on constants and factors. in: proceedings of the 3rd symposium on energy metabolism, pp. 441 – 443. K.l Blaxter (edt). london: academic press.

Bull, p., McMillan, C. and yamamoto, a., 2005. Michigan Greenhouse Gas inventory 1990 and 2002. Centre for Sustainable Systems, University of Michigan, report No. CSS05-07.

Burns, r., 2012. poultry production and climate change. xxiv World’s poultry Congress, Salvador, Brazil.


Cement and Concrete institute, 2008. Cement and Concrete Institute Review 2008. http://www.cnci.org.za/Uploads/overall_review%202008.pdf. [accessed october 2012].

CHapopSa, 2002. Charcoal potential in southern Africa. iNCo_Dev: international cooperation with developing countries (1998 – 2002).

Chhabra, a., Manjunath, K.r., panigrahy, S. and parihar, J.S., 2009. Spatial pattern of methane emissions from indian livestock. Current Science, 96 (5), 683 – 689.

Chidumayo, e.N., 1993. Responseofmiombotoharvesting:ecology and management. Stockholm environmental institute, Stockholm.

Christie, S.i. and Scholes, r.J. 1995. Carbon storage in eucalyptus and pine plantations in South africa. Environmental Monitoring and Assessment, 38: 231 – 241.

Clauss, M., Frey, r., Kiefer, B., lechner-Doll, M., loehlein, W. polster, C., rössner, G.e., Streich, W.J., 2003. the maximum attainable body size of herbivorous mammals: morphophysiological constraints on foregut , and adaptations of hindgut fermenters. oecologia, 136, 14–27.

Colgan, M.S., asner, G.p., levick, S.r., Martin, r.e. and Chadwick, o.a., 2012. topo-edaphic controls over woody plant biomass in South african savannas. Biogeosciences, 9: 1809-1821.

Cowling, r.M., richardson, D.M. and Mustart, p.J., 1997. Fynbos. in: Cowling, r.M., richardson, D.M. and pierce, S.M. (eds), Vegetation of Southern Africa. Cambridge University press, Cambridge, UK, pp 99 – 130.

Damm, o. and triebel, r., 2008. a Synthesis report on Biomass energy Consumption and availability in South africa. a report prepared for proBeC.

Department of agriculture, Forestry and Fisheries, 2010. the South african agricultural GHG inventory for 2004. Department of agriculture, Forestry and Fisheries, iSBN: 978-1-86871-321-9, DaFF, pretoria, rSa.

Department of environmental affairs (Dea), 2011: National Climate Change Response, White Paper. Department of environmental affairs, pretoria.

Department of environmental affairs (Dea), 2011. South Africa’s Second National Communication under the United Nations Framework Convention on Climate Change. Department of environmental affairs, republic of South africa, pretoria.

Department of environmental affairs (Dea), 2012. Atmospher ic Emiss ions L icenses , Department of environmental affairs, pretoria.

Department of environmental affairs (Dea), 2012. National Waste Information Baseline Report , 2012. Department of environmental affairs, pretoria.

Department of environmental affairs (Dea), 2014. South Africa’s 1st Biennial Update Report. Department of environmental affairs, pretoria. Draft, published for public comment.

Dea, 2015. National terrestrial Carbon Sinks assessment, Department of environmental affairs, pretoria.

Department of environmental affairs and tourism (Deat), 2006. South africa environment outlook. a report on the state of the environment. executive summaryandkeyfindings.DepartmentofEnvironmentalaffairs and tourism, pretoria. 42pp.

Department of environment affairs and tourism (Deat), 2009. GreenhouseGas Inventory South Africa, 1990 to2000. National inventory report, May 2009. pretoria: Department of environmental affairs and tourism (Deat, 2009).

Department of energy (Doe), 2000. DigestofSouthAfricanEnergy Statistics. pretoria.

Department of Minerals and resources, 2001a. Department ofMinerals and Resources Annual Report2001/2002. pretoria

7. references


Department of Minerals and resources, 2001b. South Africa’smineralindustry(SAMI)2001/2002. pretoria

Department of Minerals and resources, 2004. South Africa’smineralindustry(SAMI)2004/2005. pretoria



Department of Minerals and resources, 2006. South African Ferroalloy Production Trends 1995-2004. Directorate: Mineral economics, report 52/2006, pretoria.




Department of Minerals and resources, 2010c. Lime IndustrySA2010. Directorate: Mineral economics, report 85, pretoria.

Department of Minerals and resources, 2010d. Operating andDevelopingCoalMinesintheRepublicofSouthAfrica2010. Directorate: Mineral economics, Directory D2/2010, pretoria.

Department of Minerals and resources, 2010e. Producers of IndustrialMineralCommoditiesinSouthAfrica. Directorate: Mineral economics, Directory D11/2010, pretoria.

Department of Minerals and resources, 2010f. South Africa’smineralindustry(SAMI)2009/2010. pretoria

Department of Minerals and resources, 2010g. Statistical Tables, 1988 -2009. Directorate: Mineral economics, Bulletin B1/2010, pretoria.

Department of Minerals and resources, 2010h. Dolomitesand Limestone in South Africa: Supply and Demand. Directorate: Mineral economics, report r49/2005, pretoria.

Department of Minerals and resources, 2011. Operating Mines and Quarries and Mineral Processing Plants in the Republic of South Africa, 2011. Directorate: Mineral economics, Directory D1/2011, pretoria

Department of transport 2001: Statistics, pretoria. http://www.dot.gov.za [accessed october 2012].

Department of transport, 2009. Transport Statistics 2008. annual transport Statistics Bulletin, pretoria. http://www.dot.gov.za, [accessed 11 July 2012].

Department of Water affairs (DWa), 2009. Directorate: resource protection and Waste. inventory of Waste Water treatment Works, Department of Water affairs, pretoria, rSa.

Desjardins, r., rochette, p., pattey, e. and Macpherson, i., 1993. Measurements of Greenhouse Gas Fluxes using aircraft and tower-based techniques. in: Harper, l.a., Mosier, a.r. and Duxbury, J.M. (eds), agricultural ecosystems effects on trace gases and global climate change: proceedings of a symposium / sponsored by Divisions a-3 and S-3 of the american Society of agronomy and Soil Science Society of america. aSa Special publication Number 55, USa.

Dingaan, M.N.v and du preez, p.J., 2013. Grassland communities of urban open spaces in Bloemfontein, Free State, South africa. Koedoe 55: 1-8, http://www.koedoe.co.za, doi:10.4102/koedoe.v55i1.1075.

Department of Minerals and energy (DMe), 2000. HydrocarbonsandEnergyPlanningBranch–LiquidFuel, Department of Minerals and energy, pretoria.

Department of Minerals and energy (DMe), 2001. National ElectrificationProgrammeNEP. Summary evaluation report, Department of Minerals and energy, pretoria.


Department of Minerals and energy (DMe), 2002. South AfricanNationalEnergyDatabase:EnergyPrices, Department of Minerals and energy, pretoria.

Department of Minerals and energy (DMe), 2008. National EfficiencyStrategyoftheRepublicofSouthAfrica, report 32342, Department of Minerals and energy, pretoria.

Department of Minerals and energy (DMe), 2010a. Department of Minerals and Energy annual reports2009/2010, Department of Minerals and energy, pretoria.

Department of Minerals and energy (DMe), 2010b. Department of Minerals and Energy annual reports2009/2010, Department of Minerals and energy, pretoria.

Department of Mineral resources (DMr), 2010. DepartmentofMineralResourcesLimeIndustry inSouthAfrica Report, R85/2010. Department of Mineral resources, pretoria.

Department of Minerals and energy (DMe), eskom, energy research institute and University of Cape town 2002: Energy Outlook for South Africa 2002. http://www.info.gov.za/view/DownloadFileaction [accessed august 2012].

Department of energy (Doe), 2005. DigestofSouthAfricanEnergy Statistics, Department of energy, pretoria.

Department of energy (Doe), 2005. Energy Price Report. Directorate:EnergyPlanningandDevelopment, Department of energy, pretoria.

Department of energy (Doe), 2007. Biofuels IndustrialStrategyoftheRepublicofSouthAfrica, Department of Energy,Pretoria.http://www.energy.gov.za/files/esources/petroleum/biofuels_indus_strat.pdf(2).pdf [accessed october 2012].

Department of energy (Doe), 2008. Energy Balances Statistics. Department of energy, http://www.energy.gov.za/files/energyStats_frame.html.[AccessedNovember2012].

Department of energy (Doe), 2009a. DigestofSouthAfrican Energy Statistics, Department of energy, pretoria.

Department of energy (Doe), 2009b. Electrif ication Statistics. electrif ication policy Development and Management Directorate, Department of energy, pretoria.

Department of energy (Doe), 2010. South africa National energy Balance (2000-2010). Department of Energy,Pretoria,DOE.http://www.energy.gov.za/files/energyStats_frame.html

Department of energy (Doe), 2010. South African Energy Synopsis 2010. Department of energy, http://www.energy.gov.za [accessed June 2012].

Donaldson, r.a., 2009. Season effects on the potential biomass and sucrose accumulation of some commercial cultivars of sugarcane, phD thesis, University of KwaZulu-Natal, pietermaritzburg, rSa.

Donaldson, r.a., redshaw, K.a., rhodes, r. and van antwerpen, r., 2008. Season effects on productivity of some commercial South african sugarcane cultivars, 1:Biomassandradiationuseefficiency.Proc. S Afr. Sug. Technol. Ass. 81: 517 – 527.

Dovey, S.B. and Smith, C.W., 2005. Simple ratios to estimate bark and branch biomass and nutrient contents of commercial plantations. iCFr Bulletin Series 13/2005, iCFr, pietermaritzburg, rSa.

Du toit, C.J.l., Meissner, H.H. and van Niekerk, W.a., 2013a. Direct methane and nitrous oxide emissions of South african dairy and beef cattle. South African Journal of Animal Science, 43 (3): 320 – 339.

Du toit, C.J.l., Meissner, H.H. and van Niekerk, W.a., 2013d. Direct greenhouse gas emissions of the game industry in South africa. South African Journal of Animal Science, 43 (3): 376 – 393.

7. references


Du toit, C.J.l., van Niekerk, W.a. and Meissner, H.H., 2013b. Direct greenhouse gas emissions of the South african small stock sectors. South African Journal of Animal Science, 43 (3): 340 – 361.

Du toit, C.J.l., van Niekerk, W.a. and Meissner, H.H., 2013c. Direct methane and nitrous oxide emissions of monogastric livestock in South africa. South African Journal of Animal Science, 43 (3): 362 - 375.

engelbrecht, a., Golding, a., Hietkamp, S. and Scholes, r.J., 2004. ThepotentialforsequestrationofcarbondioxideinSouth Africa.CouncilforScientificandIndustrialResearch(CSir), pretoria, rSa.

environment Canada, 2009. Municipal Waste Water sources and characteristics. http://www.atl.ec.gc.ca/epb/issues/wstewtr.html

eskom 2005: Eskom Annual Report, 2005. http://www.eskom.co.za/c/article/289/publications/ [accessed February 2012]





FaoStat, 2012. Food and agriculture organization of the United Nations: Statistics. 2012. http://faostat.fao.org/site/605/default.aspx#ancor; accessed on 11/2012.

Fertilizer Society of Sa (FSSa), 2009. Fertilizer Society of South africa, http://www.fssa.org.za.

Ford, Dave, 2013. personal communication with Dave Ford, executive Director, Sa Feedlot association. telephone +27 12 667 1189, pretoria.

Forestry South africa, 2012. Forestry and Fp industry Facts, 1980 – 2011. Forestry South africa.

Garcia-apaza, e., paz, o. and arana, i., 2008. Greenhouse gas emissions from enteric fermentation of livestock in Bolivia: values for 1990–2000 and future projections. Australian Journal of Experimental Agriculture, 48: 255–259.

GCiS, 2008. South African Year Book on Transport 2007/2008. Government Communication and information System, http://www.gcis.gov.za/content/resource-centre/sa-info/yearbook/2009-10. [accessed November 2012]

GCiS, 2009. South African Year Book on Economy 2008/2009. Government Communication and information System, http://www.gcis.gov.za/content/resource-centre/sa-info/yearbook/2009-10. [accessed November 2012]

GCiS, 2010. South African Year Book on Energy 2009/2010. Government Communication and information System, http://www.gcis.gov.za/content/resource-centre/sa-info/yearbook/2009-10http://www.gcis.gov.za. [accessed November 2012]

Geoterraimage (Gti) pty ltd., 2015. 1990 – 2013-14 South african National land-cover change. Department of environmental affairs, pretoria.

Geoterraimage (Gti) pty ltd., 2015. 2013-2014 South african National land-cover dataset. Department of environmental affairs, pretoria.

Global Forest resource assessment Sa (GFraSa), 2005. GlobalForestResourcesAssessment:CountryReports–SouthAfrica. Fao, rome, italy.

Global Forest resource assessment Sa (GFraSa), 2010. GlobalForestResourcesAssessment:CountryReports–SouthAfrica. Fao, rome, italy.


Gonzalez-avalos, e. and ruiz-Suarez, l.G., 2001. Methane emission factors from cattle manure in Mexico. Biosecure technology, 80(1): 63 – 71.

Government Communication and information System (GCiS), 2009. 2009/10 Year South African Year Book. http://www.gcis.gov.za/content/resource-centre/sa-info/yearbook/2009-10 [accessed: February 2013].

Hély, C., Dowty, p.r., alleane, S., Caylor, K.K., Korontzi, S., Swap, r.J., Shugart, H.H. and Justice, C.o., 2003. regional fuel load for two climatically contrasting years in southern africa. Journal of Geophysical Research, 108.

Howden, S.M., White, D.H., McKeon, G.M., Scanlan, J.C and Carter J.o., 1994. Methods for exploring Management options to reduce Greenhouse Gas emissions from tropical Grazing Systems, Climatic Change, vol 27 pp 49–70.

Hurly, K.M., lynsky, r.J. and Stranack, r.a., 2003. protecting the privilege of Burning Sugarcane at Harvest le international Farm Management Congress http://ageconsearch.umn.edu/bitstream/24381/1/ cp03hu02.pdf

international energy agency (iea), 2012. Co2 emissions from fuel combustion: Highlights. (http://www.iea.org/co2highlights/co2highlights.pdf)

ipCC, 1997. Greenhouse Gas inventory reporting instructions. revised 1996 ipCC Guidelines for National Greenhouse Gas inventories. UNep, WMo, oeCD and iea, Bracknell United Kingdom.

ipCC, 2001. IPCCThirdAssessmentReport(TAR):ClimateChange 2001. Cambridge: intergovernmental panel on Climate Change. Cambridge University press.

ipCC, 2003. Good practice Guidance for land Use, land-Use Change and Forestry.

ipCC, 2006. 2006IPCCGuidelinesforNationalGreenhouseGasInventories. the National Greenhouse Gas inventories programme, eggleston H S, Buenida l, Miwa K, Ngara t, and tanabe K, eds; institute for Global environmental Strategies (iGeS). Hayama, Kanagawa, Japan.

ipCC, 2014. 2013 revised Supplementary Methods and Good practice Guidance arising from the Kyoto protocol, Hiraishi, t., Krug, t., tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and troxler, t.G. (eds), published: ipCC, Switzerland.

Korontzi, S., Justice, C.o. and Scholes, r.J., 2003. InfluenceoftimingandspatialextentofSavannafiresinSouthern africa on atmospheric emissions. Journal of Arid Environments, 54: 395 - 404.

Kurihara, M., Magner, t., Hunter, r.a. and McCrabb, G.J., 1999. Methane production and energy partition of cattle in the tropics. British J. Nutr. 81, 227 – 234.

laCto data, 2010. Milk producers statistics, Mpo, pretoria. vol 14 (1).

lambers, H., raven, J.a., Shaver, G.r. and Smith, S.e., 2008. plant nutrient-acquisition strategies change with soil age. Trends Ecol Evol 23: 95–103.

lechmere-oertel, r.G., 2003. the effects of goat browsing on ecosystem patterns and processes in succulent thicket, South africa. phD thesis, Nelson Mandela Metropolitan University.

lloyd p.J.D., 2001. the South african petroleum industry: a review. prepared for the automobile association of South africa. energy research institute: University of Cape town.

low, a.B. and rebelo, a.G., 1996. vegetation of South africa, lesotho and Swaziland, Department of environmental affairs and tourism, pretoria. 0-621-17316-9 iSBN.

Malimbwi, r.e. and Zahabu, e., 2009. the analysis of sustainable charcoal production systems in tanzania. in: Simmone, S., remedio, e. and trossero, M.a. (eds), Criteria and indicators for Sustainable Woodfuels: Case studies from Brazil, Guyana, Nepal, philippines and tanzania. Fao, rome, italy.

7. references


Madubansi, M., Shackleton C.M., 2007. Changes in fuelwood use and selection following electrification in the Bushbuckridge lowveld South africa. Journal of environmental Management.

Matsika r, erasmus, B.F.N., twine, W.C., 2013. a tale of two villages: assessing the dynamics of fuelwood supply in communal landscapes in South africa. environ. Conserv.

McGinn, S.M., Chen, D., loh, Z., Hill, J., Beauchemin, K.a. and Denmead, o.t., 2008. Methane emissions from feedlot cattle in australia and Canada. Australian Journal of Experimental Agriculture, 48: 183–185.

Midgley, J.J. and Seydack, a.H.W., 2006. No adverse signs of the effect of environmental change on tree biomass in the Knysna forest during the 1990s. S A J Science 102: 96-97.

Mills, a.J. and Cowling, r.M., 2006. rate of carbon sequestration at two thicket restoration sites in the eastern Cape, South africa. Restoration Ecology, 14 (1):38 – 49.

Mills, a.J., Birch, S.C., Stephenson, J.D. and Bailey, r.v., 2012. Carbon stocks in Fynbos, pastures and vineyards on the agulhas plain, South africa: a preliminary assessment. S A J Plant. Soil, iFirst: 1-3.

Mills, a.J., Cowling, r.M., Fey, M.v., Kerley, G.i.H., Donaldson, J.S., lechmere-oertel, r.G., Sigwela, a.M., Skowno, a.l. and rundel, p., 2005. effects of goat pastoralism on ecosystem carbon storage in semiarid thicket, easern Cape, South africa. Austral Ecology 30: 797 – 804.

Milton, S.J., 1990. above-ground biomass and plant cover in a succulent shrubland in the southern Karoo, South africa. S A J Bot. 56: 587-589.

Milton, S.J., yeaton, r., Dean, W.r.J. and vlok, J.H.J., 1997. Succulent Karoo. in: Cowling, r.M., richardson, D.M. and pierce, S.M. (eds.), vegetation of Southern africa. Cambridge: Cambridge University press, Cambridge, UK.

Minson, D.J. and McDonald, C.K., 1987. estimating forage intake from growth of cattle. tropical grasslands, 21(3): 116 – 121.

Moe, p.W. and tyrell, H.F., 1979. Methane production in dairy cows. J. Dairy Sci. 62: 1583 – 1586.

Moeletsi, M., paterson, G. van Der Walt, M., Malherbe, J., Nape, M., Nkambule, v. and Malaka, S., 2013. arC Contribution to alU GHG project , arC-iSCW, unpublished report no. GW/a/2013/19, pretoria, rSa.

Mucina, l. and rutherford, M.C. (eds), 2006. the vegetation of South africa, lesotho and Swaziland. Strelitzia 19. South african National Biodiversity institute, pretoria.

National electricity regulator, 2000. Electricity Supply Statistics for South Africa. http://www.ner.org.za. [accessed July 2012].



NerSa, National energy regulator of South africa (NerSa), 2006. 2006 Electricity statistics for South Africa. National energy regulator of South africa, http://www.nersa.org.za [accessed February 2012]

NerSa, 2008. 2008 electricity statistics for South africa. National energy regulator of South africa, http://www.nersa.org.za [accessed February 2012]

NerSa, 2008. the national electricity supply shortage and load shedding. National energy regulator of South africa, http://www.eepublishers.co.za/article/nersa-enquiry-into-the-national-electricity-supply-shortage-and-load-shedding.html [accessed February 2012]


o’Connor, t.G. and Bredenkamp, G.J., 1997. Grassland. in: Cowling, r.M., richardson, D.M. and pierce, S.M. (eds), Vegetation of Southern Africa. Cambridge University press, Cambridge, UK.

O’Connor,T.G.,2008.Influenceoflanduseonphytomassaccumulation in Highland Sourveld grassland in the southern Drakensberg, South africa. Afr. J. Range & Forage Sci. 25: 17 – 27.

otter, l. and Scholes, M., 2000. Methane sources and sinksinaperiodicallyfloodedSouthAfricansavannah.Glob.Biogeochem.Cycles, vol 14: 97 – 111.

otter, l. B., 2011. agricultural GHG inventory: 2000 to 2009 and forward projections to 2020. Unpublished report prepared for Department of environmental affairs, pretoria.

palmer, t. and ainslie, a., 2002. South africa: Fao CountryProfile–CountryPasture/ForageResourceprofile, rome: Fao. http://www.fao.org/ag/agp/agpc/doc/counprof/southafrica/southafrica.htm ; accessed on 15/03/2013.

pFG Glass, 2010. Chemical composition of glass. http://www.pfg.co.za/all-about-glass.aspx [accessed February 2013]

prinsloo, Magda, 2014. Statistician for the South african poultry association, personal communication via email. email: [email protected]

research Unit of Creamer Media (pty) ltd, 2009. South Africa’sSteelIndustry2009. Creamer media, Johannesburg.

roy, D.p. and Boschetti, l., 2009. Southern africa validation of the MoDiS, l3JrC and GlobCarbon Burned-area products, IEEETransactionsonGeoscienceandRemoteSensing, 47(4):1–13.

roy, D.p., Jin, y., lewis, p.e. and Justice, C.o., 2005. Prototypingaglobalalgorithmforsystematicfire-affectedarea mapping using MoDiS time series data. Remote Sensing of Environment 97(2), pp.137-162

rutherford, M.C., 1982. aboveground biomass categories of woody plants in a Burkea africana - ochna pulchra Savanna. Bothalia, 14: 131-138.

Sa Feedlot association, 2013. http://www.safeedlot.co.za/index.asp?Content=90 [accessed 10 June 2013]

Scholes, r.J. and Walker, B.H., 1993. An African Savanna: Synthesis of the Nylsvley study. Cambridge University press, Cambridge, UK.

Scholes r.J. and Hall D.o., 1996. the carbon budget of tropical savannas, woodlands and grasslands. in: Breymeyer a.i., Hall D.o., Melillo J.M. and agren G.i. (eds.) GlobalChange: Effects on Coniferous Forests and Grasslands, SCope volume 56. Wiley, Chichester. pp. 69-100.

Shackleton, C.M. and Scholes, r.J., 2011. above ground woody community attributes, biomass and carbon stocks along a rainfall gradient in the savannas of the central lowveld, South africa. S a J Bot 77: 184 -192.

Shackleton, C.M. & Shackleton, S.e. 2004.the importance of non-timber forest products in rural livelihood security and as safety-nets: evidence from South africa. South african Journal of Science, 100: 658-664.

Siebrits, F., 2012. report to authors on pig diets in South africa. pretoria.

Simalenga, t.e., Joubert, a.B.D. and Hanekom, D., 2002. agricultural Mechanization in South africa: Status, trend and Challenges, arC-all, pretoria.

Singh, J.S. and Joshi, M.C., 1979. primary production. Chapter 17 in: Coupland, r.t. (ed), Grassland ecosystems of the world: Analysis of grasslands and their uses. Cambridge University press, Cambridge, UK.

Snyman,H.A.,2005.Influenceoffireonrootdistribution,seasonal root production and root/shoot ratios in grass species in a semi-arid grassland of South africa. S A J Bot. 71: 133-144.

7. references


Snyman, H.a., 2011. root studies in a semi-arid climate. in: Grassroots, the Grassland Society of Southern africa, http://www.grassland.org.za.

Snyman, H.G., Herselman, J.e. and Kasselman, G., 2004. A metal content survey of South African Sewage Sludge and anEvaluationofAnalyticalMethodsfortheirDeterminationin Sludge. WrC report No 1283/1/04. Water research Commission, pretoria, rSa.

South africa iron and Steel institute (SaiSi), 2009. Local SalesofIronandCrudeSteelProductioninSouthAfrica. http://www.saisi.co.za/localsales.php. [accessed September 2012].

South africa iron and Steel institute (SaiSi), 2009. Local SalesofIronandCrudeSteelProductioninSouthAfrica. http://www.saisi.co.za/localsales.php. [accessed September 2012].

South africa iron and Steel institute (SaiSi), 2012. Ironand Crude Steel Production in South Africa. http://www.saisi.co.za/ironsteelproduction.php. [accessed September 2012].

South africa iron and Steel institute (SaiSi), 2012. Overview ofIronandCrudeSteelProductioninSouthAfrica. http://www.saisi.co.za/primover.php. [accessed September 2012].

South african iron and Steel institute (SaiSi), 2008. Steel Stats. http://www.saisi.co.za/index.php/steel-stats [accessed May 2012]

South african petroleum industry association (Sapia), 2006. Sapia annual report 2006. http://www.sapia.co.za. [accessed September 2011].

South african petroleum industry association (Sapia), 2008. petrol and Diesel in South africa and the impact on air Quality. http://www.sapia.co.za. [accessed September 2011].



South african Statistics Council, 2006. South african Statistics Council annual report 2005/2006, report 174/2006. http://www.statssa.gov.za/publications/StatsCouncilannualreport0506.pdf. [accessed February 2012].

South african Statistics Council, 2007. South african Statistics Council annual report 2006/2007, report.http://www.statssa.gov.za/publications/statsdownload.asp?ppN=annualreport&SCH=4026 [accessed February 2012].

South african Statistics Council, 2011. South african Statistics Council annual report 2011/11, report 174/2011. http://www.statssa.gov.za/publications/StatsCouncilannualreport0506.pdf. [accessed February 2012].

South african Wine industry Statistics, 2012. www.sawis.co.za. paarl, South africa.

Standards South africa (SSa), 2004. SaNS 1877:2004 - a standardland-coverclassificationschemeforremote-sensing applications in South africa. edition 1. pretoria, rSa.

Statistics South africa (Statistics Sa), 2006. Statistics South Africa Annual Report 2005/2006. http://www.statssa.gov.za/annual reports. [accessed September 2012]

Statistics South africa, 2010. Mid-year population estimates: 2010. Statistical release p0302.

Statistics South africa, 2013. Mid-year population estimates: 2013. Statistical release p0302.

thompson, M.W., 1999. South african National land-cover database project. Data users manual, Final report. CSir-arC Contract report. eNv/p/C 98136, pretoria.


United Nations (UN), 2012. population Division of the Department of economic and Social affairs of the Secretariat, World population prospects: http://esa.un.org/unpd/wpp/index.htm [accessed april 2013]

van den Berg, e.C., plarre, C., van den Berg, H.M. and thompson, M.W., 2008. The South African National Land-Cover 2000. agricultural research Council-institute for Soil, Climate and Water, Unpublished report No. GW/a/2008/86.

van der Merwe, M. and Scholes, r. J., 1998. Greenhouse GasEmissionsInventoryof1990forSouthAfrica,Phase1. Council forScientificandIndustrialResearch(CSIR),pretoria: National research Foundation, rSa.

van der vyver, M.l., 2011. restoring the biodiversity of canopy species within degraded Spekboom thicket. MSc thesis, Nelson Mandela Metropolitan University.

van der Waal, C., 2008. Guidelines for the Utilization andDisposalofWastewaterSludge,Volume1-5: ImpactAssessment. WrC report No tt 370/08. Water research Commission, pretoria, rSa.

van Dyk J.p., 2006. An overview of the Zincor Process, Southern African Pyrometallurgy 2006. South african institute of Mining and Metallurgy, Johannesburg

van Heerden, p.D.r., Donaldson, r.a., Watt, D.a. and Singels, a., 2010. Biomass accumulation in sugarcane: unravelling the factors underpinning reduced growth phenomena. J Exp. Bot., June 13, 2010, p 1 – 11; doi:10.1093/jeb/erq144.

Wang, S. and Huang, D., 2005. assessment of Greenhouse Gas emissions from poultry enteric Fermentation. asian-australian Journal of Animal Science 18:873-878.

Wessels, K.J., Colgan, M.S., erasmus, B.F.N., asner, G.p., twine, W.C., Mathieu, r., aardt, J.a.N., Fisher, J.t. and Smit, i.p.J., 2013. Unsustainable fuelwood extraction from South african savannas. Environ. Res. Lett. 8: 1-10.

World Steel association, 2010. Key Facts about World Steel industry. http://www.worldsteel.org/media-centre/key-facts.html . [accessed May 2012].

7. references


APPENDICESJanuary 2017


8. APPENDIx A: SUMMary taBleS

INVENTORY YEAR: 2000 Emissions/removals (Gg Co2eq)

Categories Net CO2 CH4 N2O HFCs PFCs Total

Total including FOLU 357 791.87 48 367.06 27 162.82 0.00 982.24 434 303.98

Total excluding FOLU 367 050.44 47 937.81 27 162.82 0.00 982.24 443 133.31

1 - Energy 335 519.08 5 002.45 2 070.76 342 592.29

1.a - Fuel Combustion activities 306 596.79 572.96 2 070.76 309 240.52

1.A.1-EnergyIndustries 219 524.69 58.12 963.57 220 546.38

1.A.2-ManufacturingIndustriesandConstruction 32 505.42 8.93 138.17 32 652.52

1.A.3 - Transport 35 214.80 270.27 531.12 36 016.18

1.A.4 - Other Sectors 18 366.30 234.66 435.38 19 036.34

1.A.5-Non-Specified 985.58 0.98 2.53 989.09

1.B - Fugitive emissions from fuels 28 922.29 4 429.49 33 351.77

1.B.1 - Solid Fuels 23.67 1 978.88 2 002.55

1.B.2 - Oil and Natural Gas 752.04 752.04

1.B.3 - Other emissions from Energy Production 28 146.58 2 450.60 30 597.19

2 - Industrial Processes and Product Use 30 935.81 75.50 1 570.25 982.24 33 563.79

2.a - Mineral industry 3 847.79 3 847.79

2.A.1 - Cement production 3 347.05 3 347.05

2.A.2 - Lime production 426.37 426.37

2.A.3 - Glass Production 74.36 74.36

2.B - Chemical industry 1 063.45 71.88 1 570.25 2 705.58

2.C - Metal industry 25 828.66 3.62 982.24 26 814.51

2.C.1-IronandSteelProduction 16 410.53 16 410.53

2.C.2 - Ferroalloys Production 8 079.14 3.62 8 082.76

2.C.3 - Aluminium production 1 105.47 982.24 2 087.71

2.C.4 - Magnesium production

2.C.5 - Lead Production 39.16 39.16

2.C.6 - Zinc Production 194.36 194.36

2.D - Non-energy products from Fuels and Solvent Use

195.92 195.92

2.D.1-LubricantUse 188.48 188.48

2.D.2-ParaffinWaxUse 7.44 7.44

2.F - product Uses as Substitutes for oDS

3 - Agriculture, Forestry, and Other Land Use -7 702.83 31 573.24 22 949.57 46 819.97

3.a - livestock 30 155.07 1 007.32 31 162.38

3.A.1 - Enteric Fermentation 29 290.35 29 290.35

3.A.2 - Manure Management 864.71 1 007.32 1 872.03

8. appendix a




3.B - land -7 986.19 429.25 -7 556.94

3.B.1 - Forest land -21 565.09 -21 565.09

3.B.2 - Cropland 4 480.05 4 480.05

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 429.25 429.25

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20

3.C - aggregate sources and non-Co2 emissions sources on land

595.55 988.92 21 942.25 23 526.72

3.C.1-Emissionsfrombiomassburning 988.92 914.31 1 903.24

3.C.2 - Liming 384.05 384.05

3.C.3 - Urea application 211.49 211.49

3.C.4-DirectN2O Emissions from managed soils 15 909.52 15 909.52

3.C.5-IndirectN2O Emissions from managed soils 4 768.44 4 768.44

3.C.6-IndirectN2O Emissions from manure management

349.97 349.97

3.D - other -312.19 -312.19

3.D.1-HarvestedWoodProducts -312.19 -312.19

4 - Waste 11 715.88 572.24 12 288.12

4.a - Solid Waste Disposal 9 367.55 9 367.55

4.D - Wastewater treatment and Discharge 2 348.33 572.24 2 920.57

Memo Items (5)

International bunkers 2 972.40 2.87 6.73 2 981.99

1.A.3.a.i-InternationalAviation(InternationalBunkers) 2 972.40 2.87 6.73 2 981.99

1.A.3.d.i-Internationalwater-bornenavigation(Internationalbunkers)

1.A.5.c - Multilateral Operations




Total including FOLU 356 779.41 49 243.42 27 062.21 0.00 1 008.22 434 093.27

Total excluding FOLU 366 067.83 48 827.62 27 062.21 0.00 1 008.22 442 965.89

1 - Energy 334 283.28 4 991.41 2 062.81 341 337.50




1.A.3 - Transport 35 448.67 269.24 533.67 36 251.59

1.A.4 - Other Sectors 21 825.31 234.55 449.20 22 509.06

1.A.5-Non-Specified 980.25 0.98 2.51 983.74


1.B.1 - Solid Fuels 23.35 1 966.46 1 989.81



2 - Industrial Processes and Product Use 31 140.73 75.72 1 512.39 1 008.22 32 728.83






2.C - Metal industry 25 939.17 3.56 1 008.22 25 942.73



2.C.3 - Aluminium production 1 116.55 1 008.22 1 116.55





225.97 225.97





3.a - livestock 29 992.39 1 010.84 31 003.23



8. appendix a




3.B - land -7 653.70 415.80 -7 237.90

3.B.1 - Forest land -21 326.32 -21 326.32

3.B.2 - Cropland 4 573.77 4 573.77

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 415.80 415.80

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


643.82 1 146.77 21 892.11 23 682.69

3.C.1-Emissionsfrombiomassburning 1 146.77 1 089.52 2 236.29

3.C.2 - Liming 497.15 497.15





352.76 352.76

3.D - other -674.52 -674.52


4 - Waste 12 621.33 584.07 13 205.41



Memo Items (5)










1 - Energy 345 080.78 5 022.89 2 112.26 352 215.93




1.A.3 - Transport 36 050.18 269.83 540.11 36 860.12

1.A.4 - Other Sectors 24 918.15 234.20 460.84 25 613.19

1.A.5-Non-Specified 979.76 0.98 2.51 983.25


1.B.1 - Solid Fuels 23.02 1 938.08 1 961.10









2.C - Metal industry 27 537.39 4.43 898.28 27 541.82








250.32 250.32





3.a - livestock 29 743.58 1 027.60 30 771.18



8. appendix a




3.B - land -10 032.41 402.35 -9 630.06

3.B.1 - Forest land -23 798.42 -23 798.42

3.B.2 - Cropland 4 667.15 4 667.15

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 402.35 402.35

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 175.54 1 153.54 22 277.66 24 606.73


3.C.2 - Liming 683.69 683.69





358.22 358.22

3.D - other -817.15 -817.15


4 - Waste 13 496.19 592.17 14 088.37



Memo Items (5)










1 - Energy 368 291.84 5 187.23 2 238.93 375 718.00


1.A.1-EnergyIndustries 237 730.31 63.04 1 052.05 238 845.40


1.A.3 - Transport 37 825.84 279.15 558.09 38 663.08

1.A.4 - Other Sectors 27 194.38 238.28 475.71 27 908.37

1.A.5-Non-Specified 1 011.23 1.01 2.59 1 014.82


1.B.1 - Solid Fuels 24.85 2 092.96 2 117.81

1.B.2 - Oil and Natural Gas 1 457.99 1 457.99


2 - Industrial Processes and Product Use 32 850.94 75.65 936.31 896.23 33 862.90





2.B - Chemical industry 1 123.51 71.20 936.31 2 131.02

2.C - Metal industry 27 340.02 4.45 896.23 27 344.47








248.61 248.61





3.a - livestock 29 143.08 1 050.50 30 193.58



8. appendix a




3.B - land -9 282.59 388.90 -8 893.69

3.B.1 - Forest land -23 142.12 -23 142.12

3.B.2 - Cropland 4 760.67 4 760.67

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 388.90 388.90

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


917.19 850.88 21 161.27 22 929.34


3.C.2 - Liming 585.99 585.99





360.63 360.63

3.D - other -927.35 -927.35


4 - Waste 14 346.76 599.76 14 946.52



Memo Items (5)










1 - Energy 385 957.52 5 336.79 2 312.84 393 607.16




1.A.3 - Transport 39 384.78 287.96 580.39 40 253.12

1.A.4 - Other Sectors 31 957.93 238.05 487.07 32 683.05

1.A.5-Non-Specified 1 041.37 1.04 2.67 1 045.08


1.B.1 - Solid Fuels 25.43 2 141.35 2 166.78









2.C - Metal industry 27 295.53 4.63 888.40 27 300.16








246.18 246.18





3.a - livestock 28 972.91 1 066.06 30 038.97



8. appendix a




3.B - land -7 117.31 375.46 -6 741.86

3.B.1 - Forest land -21 072.76 -21 072.76

3.B.2 - Cropland 4 856.59 4 856.59

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 375.46 375.46

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 010.10 795.78 21 108.50 22 914.38


3.C.2 - Liming 585.54 585.54





364.54 364.54

3.D - other -1 185.17 -1 185.17

3.D.1-HarvestedWoodProducts -1 185.17 -1 185.17

4 - Waste 15 173.07 606.90 15 779.97



Memo Items (5)










1 - Energy 379 149.49 5 139.90 2 297.55 386 586.94




1.A.3 - Transport 40 738.29 293.55 596.39 41 628.23

1.A.4 - Other Sectors 32 109.14 227.68 475.11 32 811.92

1.A.5-Non-Specified 1 058.49 1.05 2.71 1 062.25


1.B.1 - Solid Fuels 25.60 2 155.57 2 181.17



2 - Industrial Processes and Product Use 34 473.06 103.56 1 971.93 841.90 871.87 38 262.32





2.B - Chemical industry 818.00 99.16 1 971.93 2 889.09

2.C - Metal industry 28 347.05 4.40 871.87 29 223.32








467.85 467.85



2.F - product Uses as Substitutes for oDS 841.90 841.90


3.a - livestock 28 866.77 1 131.24 29 998.01



8. appendix a




3.B - land -7 987.02 362.01 -7 625.01

3.B.1 - Forest land -22 034.70 -22 034.70

3.B.2 - Cropland 4 948.84 4 948.84

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 362.01 362.01

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


607.00 1 267.35 20 957.41 22 831.77


3.C.2 - Liming 267.41 267.41





382.32 382.32

3.D - other -197.37 -197.37


4 - Waste 15 976.98 613.83 16 590.81



Memo Items (5)








Total including FOLU 410 389.31 52 064.40 27 212.50 1 107.58 934.55 491 708.33

Total excluding FOLU 421 926.95 51 715.84 27 212.50 1 107.58 934.55 502 897.41

1 - Energy 386 029.62 4 943.90 2 102.83 393 076.35




1.A.3 - Transport 41 792.69 297.26 608.80 42 698.74

1.A.4 - Other Sectors 35 075.12 93.31 254.49 35 422.92

1.A.5-Non-Specified 1 069.30 1.06 2.74 1 073.10


1.B.1 - Solid Fuels 25.58 2 154.20 2 179.78



2 - Industrial Processes and Product Use 35 104.85 113.30 2 038.21 1 107.58 934.55 39 298.49






2.C - Metal industry 28 889.38 4.90 934.55 29 828.82








509.15 509.15



2.F - product Uses as Substitutes for oDS 1 107.58 1 107.58


3.a - livestock 28 761.12 1 194.68 29 955.79



8. appendix a




3.B - land -9 695.62 348.56 -9 347.06

3.B.1 - Forest land -23 840.67 -23 840.67

3.B.2 - Cropland 5 046.20 5 046.20

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 348.56 348.56

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


792.48 1 137.15 21 256.02 23 185.65


3.C.2 - Liming 445.96 445.96





397.84 397.84

3.D - other -881.82 -881.82


4 - Waste 16 760.37 620.76 17 381.13

4.a - Solid Waste Disposal 14 212.91 0.00 14 212.91


Memo Items (5)










1 - Energy 414 451.61 5 047.57 2 253.89 421 753.06




1.A.3 - Transport 44 230.25 306.27 643.74 45 180.26

1.A.4 - Other Sectors 36 194.68 96.03 259.70 36 550.41

1.A.5-Non-Specified 1 095.70 1.09 2.81 1 099.60


1.B.1 - Solid Fuels 25.88 2 179.14 2 205.01



2 - Industrial Processes and Product Use 34 108.74 152.84 1 195.78 1 077.88 970.11 37 505.35






2.C - Metal industry 28 076.12 4.60 970.11 29 050.83








234.08 234.08





3.a - livestock 28 088.53 1 236.93 29 325.46



8. appendix a




3.B - land -10 961.33 335.11 -10 626.22

3.B.1 - Forest land -25 206.57 -25 206.57

3.B.2 - Cropland 5 146.39 5 146.39

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 335.11 335.11

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


984.81 1 012.63 20 602.49 22 599.93

3.C.1-Emissionsfrombiomassburning 1 012.63 880.03 1 892.66

3.C.2 - Liming 524.87 524.87





411.23 411.23

3.D - other -581.39 -581.39


4 - Waste 17 524.16 627.60 18 151.76



Memo Items (5)










1 - Energy 406 097.55 5 353.88 2 246.75 413 698.19




1.A.3 - Transport 43 145.12 294.22 625.89 44 065.23

1.A.4 - Other Sectors 38 190.91 119.15 309.36 38 619.42

1.A.5-Non-Specified 1 049.34 1.04 2.69 1 053.07


1.B.1 - Solid Fuels 26.35 2 219.14 2 245.50



2 - Industrial Processes and Product Use 32 738.93 78.38 529.63 1 039.92 548.12 34 934.98





2.B - Chemical industry 574.34 73.85 529.63 1 177.83

2.C - Metal industry 26 832.68 4.52 548.12 27 385.32








221.07 221.07





3.a - livestock 28 742.71 1 226.33 29 969.04



8. appendix a




3.B - land -4 651.82 321.66 -4 330.15

3.B.1 - Forest land -18 988.89 -18 988.89

3.B.2 - Cropland 5 238.22 5 238.22

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 321.66 321.66

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 104.59 1 098.35 21 124.55 23 327.49


3.C.2 - Liming 658.92 658.92





410.32 410.32

3.D - other -780.75 -780.75


4 - Waste 18 270.10 634.57 18 904.67



Memo Items (5)










1 - Energy 414 213.63 5 333.58 2 286.63 421 833.84




1.A.3 - Transport 43 765.46 300.31 635.12 44 700.89

1.A.4 - Other Sectors 40 956.32 119.23 320.12 41 395.66

1.A.5-Non-Specified 1 072.49 1.07 2.75 1 076.30


1.B.1 - Solid Fuels 26.18 2 204.26 2 230.44



2 - Industrial Processes and Product Use 31 353.12 56.58 512.02 995.54 108.25 33 025.51






2.C - Metal industry 25 474.31 3.60 108.25 25 586.17








233.76 233.76



2.F - product Uses as Substitutes for oDS 995.54 995.54


3.a - livestock 28 174.26 1 237.97 29 412.23



8. appendix a




3.B - land -12 373.92 308.22 -12 065.71

3.B.1 - Forest land -26 788.44 -26 788.44

3.B.2 - Cropland 5 315.67 5 315.67

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 308.22 308.22

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


984.02 1 009.96 20 728.90 22 722.87


3.C.2 - Liming 616.23 616.23





412.83 412.83

3.D - other -98.32 -98.32


4 - Waste 18 997.54 641.43 19 638.96



Memo Items (5)










1 - Energy 427 401.95 5 354.59 2 360.68 435 117.22




1.A.3 - Transport 46 617.91 318.14 670.45 47 606.50

1.A.4 - Other Sectors 44 683.84 115.51 325.07 45 124.41

1.A.5-Non-Specified 1 134.86 1.13 2.91 1 138.90


1.B.1 - Solid Fuels 26.59 2 239.46 2 266.05



2 - Industrial Processes and Product Use 32 835.90 67.13 325.54 2 096.25 138.26 35 463.07






2.C - Metal industry 27 186.27 4.24 138.26 27 328.78








233.87 233.87





3.a - livestock 28 996.25 1 249.22 30 245.47



8. appendix a




3.B - land -14 210.93 294.77 -13 916.16

3.B.1 - Forest land -28 732.32 -28 732.32

3.B.2 - Cropland 5 422.53 5 422.53

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 294.77 294.77

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 070.06 1 232.70 21 274.88 23 577.63


3.C.2 - Liming 585.54 585.54





414.85 414.85

3.D - other -490.26 -490.26


4 - Waste 19 705.80 648.58 20 354.38



Memo Items (5)








Total including FOLU 426 660.77 57 071.31 26 364.51 1 749.86 2 410.85 514 257.30

Total excluding FOLU 442 653.44 56 789.99 26 364.51 1 749.86 2 410.85 529 968.65

1 - Energy 407 525.47 6 053.19 2 263.95 415 842.61




1.A.3 - Transport 47 293.87 318.80 669.94 48 282.61

1.A.4 - Other Sectors 39 157.78 111.02 292.12 39 560.92

1.A.5-Non-Specified 1 134.09 1.13 2.91 1 138.13


1.B.1 - Solid Fuels 35.05 2 951.32 2 986.37



2 - Industrial Processes and Product Use 33 991.31 92.45 643.36 1 749.86 2 410.85 38 887.83






2.C - Metal industry 28 610.64 4.52 2 410.85 31 026.01




2.C.4 - Magnesium production 0.00




196.39 268.00





3.a - livestock 29 024.32 1 324.46 30 348.78



8. appendix a




3.B - land -16 074.59 281.32 -15 793.26

3.B.1 - Forest land -29 731.24 -29 731.24

3.B.2 - Cropland 5 518.00 5 518.00

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 281.32 281.32

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 136.66 1 142.29 21 459.44 23 738.38


3.C.2 - Liming 585.54 585.54





428.77 428.77

3.D - other 81.91 81.91

3.D.1-HarvestedWoodProducts 81.91 81.91

4 - Waste 20 477.74 673.30 21 151.04



Memo Items (5)








Total including FOLU 433 839.25 55 436.76 25 645.87 1 396.12 1 979.19 518 297.18

Total excluding FOLU 454 921.51 55 168.88 25 645.87 1 396.12 1 979.19 539 111.58

1 - Energy 420 846.23 4 923.31 2 342.21 428 111.75




1.A.3 - Transport 46 713.90 312.66 663.06 47 689.62

1.A.4 - Other Sectors 40 174.62 107.05 289.22 40 570.89

1.A.5-Non-Specified 1 110.53 1.11 2.85 1 114.48


1.B.1 - Solid Fuels 20.65 1 739.17 1 759.83



2 - Industrial Processes and Product Use 32 934.73 73.55 745.00 1 396.12 1 979.19 37 128.59





2.B - Chemical industry 516.54 70.18 745.00 0.00 0.00 1 331.71

2.C - Metal industry 27 738.70 3.37 1 979.19 29 721.26


2.C.2 - Ferroalloys Production 11 623.80 3.37 0.00 0.00 0.00 11 627.18




2.C.6 - Zinc Production


268.00 268.00





3.a - livestock 28 050.11 1 362.80 29 412.91



8. appendix a




3.B - land -20 570.44 267.87 -20 302.57

3.B.1 - Forest land -34 325.17 -34 325.17

3.B.2 - Cropland 5 616.07 5 616.07

3.B.3 - Grassland 7 904.13 7 904.13

3.B.4 - Wetlands 267.87 267.87

3.B.5 - Settlements 1 194.72 1 194.72

3.B.6 - Other lands -960.20 -960.20


1 140.55 882.25 20 506.82 22 529.61


3.C.2 - Liming 585.54 585.54





436.87 436.87

3.D - other -511.83 -511.83


4 - Waste 21 239.67 689.05 21 928.72



Memo Items (5)






9. APPENDIx b: Key CateGory aNalySiS

IPCC Category

codeIPCC Category GHG

2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


Co2 185027.4 41.3 41.3

1.a.3.b transport - liquid Fuels - road transportation Co2 32623.3 7.3 48.6


Co2 30454.7 6.8 55.4

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 29101.6 6.5 61.9

3.a.1.a enteric Fermentation - Cattle CH4 22819.4 5.1 67.0

2.C.1 Metal industry - iron and Steel production Co2 16410.5 3.7 70.7

3.B.1.b Forest land - land Converted to Forest land Co2 -15339.9 3.4 74.1


N2o 13180.8 2.9 77.0

4.a Solid Waste Disposal CH4 9367.6 2.1 79.1

2.C.2 Metal industry - Ferroalloys production Co2 8079.1 1.8 80.9

3.B.3.b Grassland - land Converted to Grassland Co2 7904.1 1.8 82.7

1.a.4.a other Sectors - liquid Fuels - Commercial/institutional Co2 7690.4 1.7 84.4

3.B.1.a Forest land - Forest land remaining Forest land Co2 -6225.2 1.4 85.8

3.B.2.b Cropland - land Converted to Cropland Co2 5721.7 1.3 87.1

3.a.1.c enteric Fermentation - Sheep CH4 4169.0 0.9 88.0

1.a.4.b other Sectors - Solid Fuels - residential Co2 3604.2 0.8 88.8

2.a.1 Mineral industry - Cement production Co2 3347.1 0.7 89.5

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining Co2 3164.8 0.7 90.3

1.a.4.b other Sectors - liquid Fuels - residential Co2 2868.9 0.6 90.9


N2o 2479.0 0.6 91.4

4.D.1 Wastewater treatment and Discharge - Domestic CH4 2348.3 0.5 92.0


Co2 2217.7 0.5 92.5


Co2 2207.2 0.5 93.0

9.1 Level assessment: 2000 (including FolU)

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)

1.a.3.a transport - liquid Fuels - Civil aviation Co2 2040.0 0.5 93.4


N2o 1939.5 0.4 93.8


N2o 1934.7 0.4 94.3

1.a.4.a other Sectors - Solid Fuels - Commercial/institutional Co2 1811.2 0.4 94.7

2.B Chemical industry - industry F N2o Ca 0.4 95.0

3.B.2.a Cropland - Cropland remaining Cropland Co2 -1241.6 0.3 95.3

3.B.5.b Settlements - land Converted to Settlements Co2 1191.0 0.3 95.6


Co2 1186.1 0.3 95.8

3.a.1.j enteric Fermentation - other CH4 1106.7 0.2 96.1

2.C.3 Metal industry - aluminium production Co2 1105.5 0.2 96.3

3.a.1.d enteric Fermentation - Goats CH4 993.6 0.2 96.6

1.a.5.a Non-Specified-LiquidFuels-Stationary Co2 985.6 0.2 96.8

2.C.3 Metal industry - aluminium production pFCs 982.2 0.2 97.0

3.B.6.b other land - land Converted to other land Co2 -960.2 0.2 97.2


N2o 853.5 0.2 97.4

3.a.2.i Manure management - poultry N2o 642.6 0.1 97.5

3.C.1.c emissions from biomass burning - Grasslands N2o 637.5 0.1 97.7

3.a.2.h Manure management - Swine CH4 612.1 0.1 97.8

4.D.1 Wastewater treatment and Discharge - Domestic N2o 572.2 0.1 98.0

1.a.3.c transport - liquid Fuels - railways Co2 551.5 0.1 98.1

3.C.1.c emissions from biomass burning - Grasslands CH4 542.5 0.1 98.2

2.B Chemical industry - industry x Co2 Ca 0.1 98.3

1.a.3.b transport - liquid Fuels - road transportation N2o 463.0 0.1 98.4

3.C.4Direct N2o emissions from managed soils - organic N application

N2o 441.3 0.1 98.5

2.B Chemical industry - industry K Co2 Ca 0.1 98.6


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)

3.B.4.a. Wetlands-Floodedlandremainingfloodedland CH4 429.2 0.1 98.7

2.a.2 Mineral industry - lime production Co2 426.4 0.1 98.8

3.C.2 liming Co2 384.1 0.1 98.9

1.a.4.b other Sectors - Biomass - residential N2o 380.9 0.1 99.0

3.C.4Direct N2o emissions from managed soils - N in crop residues

N2o 352.8 0.1 99.0

3.C.6 indirect N2o emissions from manure management N2o 350.0 0.1 99.1

3.D.1 Harvested Wood products Co2 -312.2 0.1 99.2

3.a.2.a Manure management - Cattle N2o 301.0 0.1 99.3

1.a.3.b transport - liquid Fuels - road transportation CH4 267.6 0.1 99.3

3.C.1.b emissions from biomass burning - Croplands CH4 244.2 0.1 99.4

1.a.4.b other Sectors - Biomass - residential CH4 222.0 0.0 99.4

3.C.3 Urea application Co2 211.5 0.0 99.5

2.C.6 Metal industry - Zinc production Co2 194.4 0.0 99.5

2.D.1Non-energy products from Fuels and Solvent Use - lubricant use

Co2 188.5 0.0 99.6

3.a.2.a Manure management - Cattle CH4 182.6 0.0 99.6

1.a.4.cother Sectors - Solid Fuels - agriculture/Forestry/Fishing/Fish farms

Co2 171.5 0.0 99.6

3.C.1.a emissions from biomass burning - Forest lands CH4 164.6 0.0 99.7

3.C.1.a emissions from biomass burning - Forest lands N2o 151.1 0.0 99.7

2.B Chemical industry - industry Q Co2 Ca 0.0 99.7

1.a.2 Manufacturing industries and Construction - Solid Fuels N2o 134.2 0.0 99.8

3.a.1.f enteric Fermentation - Horses CH4 111.8 0.0 99.8


N2o 105.4 0.0 99.8

3.C.1.b emissions from biomass burning - Croplands N2o 81.5 0.0 99.8

2.a.3 Mineral industry - Glass production Co2 74.4 0.0 99.9

2.B Chemical industry - industry x CH4 Ca 0.0 99.9

3.a.2.h Manure management - Swine N2o 63.7 0.0 99.9

1.a.3.c transport - liquid Fuels - railways N2o 63.0 0.0 99.9

3.a.2.i Manure management - poultry CH4 54.2 0.0 99.9

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)

3.a.1.h enteric Fermentation - Swine CH4 52.1 0.0 99.9


CH4 44.2 0.0 99.9

2.C.5 Metal industry - lead production Co2 39.2 0.0 99.9

3.a.1.g enteric Fermentation - Mules and assess CH4 37.7 0.0 99.9

3.C.1.e emissions from biomass burning - other lands N2o 23.1 0.0 100.0

3.C.1.d emissions from biomass burning - Wetlands N2o 21.1 0.0 100.0

3.C.1.e emissions from biomass burning - other lands CH4 19.7 0.0 100.0

1.a.4.a other Sectors - liquid Fuels - Commercial/institutional N2o 18.3 0.0 100.0

3.C.1.d emissions from biomass burning - Wetlands CH4 18.0 0.0 100.0

1.a.4.b other Sectors - Solid Fuels - residential N2o 16.6 0.0 100.0

1.a.4.aother Sectors - Gaseous Fuels - Commercial/institutional

Co2 13.0 0.0 100.0

1.a.1.cenergy industries - liquid Fuels - Manufacture of spolid fuels and other energy industries

CH4 11.6 0.0 100.0

3.a.2.f Manure management - Horses CH4 10.2 0.0 100.0

1.a.4.a other Sectors - Solid Fuels - Commercial/institutional N2o 8.3 0.0 100.0

2.D.1Non-energy products from Fuels and Solvent Use - Paraffinwaxuse

Co2 7.4 0.0 100.0

1.a.4.a other Sectors - liquid Fuels - Commercial/institutional CH4 7.1 0.0 100.0

1.a.2 Manufacturing industries and Construction - Solid Fuels CH4 7.0 0.0 100.0


N2o 5.3 0.0 100.0

1.a.4.b other Sectors - liquid Fuels - residential N2o 5.2 0.0 100.0

1.a.3.a transport - liquid Fuels - Civil aviation N2o 5.1 0.0 100.0

2.C.2 Metal industry - Ferroalloys production CH4 3.6 0.0 100.0

3.a.2.g Manure management - Mules and asses CH4 3.4 0.0 100.0

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining N2o 3.2 0.0 100.0


N2o 2.8 0.0 100.0

aConfidential-disaggregatedChemicalIndustry[2B]dataisnotshown.


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


Co2 185027.4 45.3 45.3

1.a.3.b transport - liquid Fuels - road transportation Co2 32623.3 8.0 53.3


Co2 30454.7 7.5 60.7

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 29101.6 7.1 67.9

3.a.1.a enteric Fermentation - Cattle CH4 22819.4 5.6 73.4

2.C.1 Metal industry - iron and Steel production Co2 16410.5 4.0 77.5


N2o 13180.8 3.2 80.7

4.a Solid Waste Disposal CH4 9367.6 2.3 83.0

2.C.2 Metal industry - Ferroalloys production Co2 8079.1 2.0 84.9

1.a.4.a other Sectors - liquid Fuels - Commercial/institutional Co2 7690.4 1.9 86.8

3.a.1.c enteric Fermentation - Sheep CH4 4169.0 1.0 87.9

1.a.4.b other Sectors - Solid Fuels - residential Co2 3604.2 0.9 88.7

2.a.1 Mineral industry - Cement production Co2 3347.1 0.8 89.6

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining Co2 3164.8 0.8 90.3

1.a.4.b other Sectors - liquid Fuels - residential Co2 2868.9 0.7 91.0


N2o 2479.0 0.6 91.6

4.D.1 Wastewater treatment and Discharge - Domestic CH4 2348.3 0.6 92.2


Co2 2217.7 0.5 92.8


Co2 2207.2 0.5 93.3

1.a.3.a transport - liquid Fuels - Civil aviation Co2 2040.0 0.5 93.8


N2o 1939.5 0.5 94.3


N2o 1934.7 0.5 94.7

1.a.4.a other Sectors - Solid Fuels - Commercial/institutional Co2 1811.2 0.4 95.2

9.2 Level assessment: 2000 (excluding FolU)

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)



Co2 1186.1 0.3 95.9







N2o 853.5 0.2 97.3










N2o 441.3 0.1 98.5



3.C.2 liming Co2 384.1 0.1 98.9



N2o 352.8 0.1 99.0








IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)

2.C.6 Metal industry - Zinc production Co2 194.4 0.0 99.5


Co2 188.5 0.0 99.5



Co2 171.5 0.0 99.6







N2o 105.4 0.0 99.8









CH4 44.2 0.0 99.9










Co2 13.0 0.0 100.0

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


CH4 11.6 0.0 100.0

3.a.2.f Manure management – Horses CH4 10.2 0.0 100.0



Co2 7.4 0.0 100.0




N2o 5.3 0.0 100.0

1.a.4.b other Sectors - liquid Fuels – residential N2o 5.2 0.0 100.0




1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining N2o 3.2 0.0 100.0



IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


Co2 242 154.0 43.5 43.5

1.a.3.b transport - liquid Fuels - road transportation Co2 42 929.6 7.7 51.2


Co2 29 903.3 5.4 56.6

3.B.1.b Forest land - land Converted to Forest land Co2 -29 490.8 5.3 61.9

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 24 472.7 4.4 66.3





1.a.4.a other Sectors - liquid Fuels - Commercial/institutional Co2 12 685.4 2.3 81.7


N2o 11 815.0 2.1 83.9


3.B.3.b Grassland - land Converted to Grassland Co2 7 871.8 1.4 87.4

3.B.2.b Cropland - land Converted to Cropland Co2 5 721.0 1.0 88.4

3.B.1.a Forest land - Forest land remaining Forest land Co2 -4 834.4 0.9 89.3




1.a.4.a other Sectors - Solid Fuels - Commercial/institutional Co2 3 434.8 0.6 91.9


Co2 3 420.3 0.6 92.5

4.D.1 Wastewater treatment and Discharge - Domestic CH4 2 827.7 0.5 93.0

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining Co2 2 802.4 0.5 93.5


Co2 2 497.5 0.4 93.9


N2o 2 330.0 0.4 94.4


9.3 Level assessment: 2012 (including FolU)

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


Co2 2 114.0 0.4 95.1


N2o 2000.1 0.4 95.5



N2o 1791.7 0.3 96.2


HFCs 1396.1 0.3 96.4

3.B.5.b Settlements - land Converted to Settlements Co2 1183.2 0.2 96.6


N2o 1117.1 0.2 96.8




3.B.6.b other land - land Converted to other land Co2 -960.3 0.2 97.6







3.C.2 liming Co2 585.5 0.1 98.5



3.D.1 Harvested Wood products Co2 -511.8 0.1 98.8


N2o 489.8 0.1 98.9





N2o 396.9 0.1 99.2


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)




3.B.4.a. Wetlands-Floodedlandremainingfloodedland CH4 267.9 0.0 99.4


Co2 250.6 0.0 99.5








N2o 141.5 0.0 99.7






3.B.2.a Cropland - Cropland remaining Cropland Co2 -104.9 0.0 99.8









CH4 57.9 0.0 99.9

3.a.1.h enteric Fermentation – Swine CH4 50.0 0.0 99.9



9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)

1.a.3.c transport - liquid Fuels – railways N2o 36.3 0.0 100.0




Co2 17.6 0.0 100.0


Co2 17.4 0.0 100.0


3.C.1.d emissions from biomass burning – Wetlands CH4 15.6 0.0 100.0

3.C.1.d emissions from biomass burning – Wetlands N2o 12.5 0.0 100.0


3.a.2.f Manure management – Horses CH4 11.6 0.0 100.0



CH4 8.4 0.0 100.0


N2o 8.2 0.0 100.0


2.B Chemical industry - industry S Co2 Ca 0.0 100.0


N2o 4.9 0.0 100.0

1.a.4.b other Sectors - Solid Fuels – residential CH4 4.4 0.0 100.0





IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)


Co2 242 154.0 47.9 47.9

1.a.3.b transport - liquid Fuels - road transportation Co2 42 929.6 8.5 56.4


Co2 29 903.3 5.9 62.3

1.a.2 Manufacturing industries and Construction - Solid Fuels Co2 24 472.7 4.8 67.2





1.a.4.a other Sectors - liquid Fuels - Commercial/institutional Co2 12 685.4 2.5 84.1


N2o 11 815.0 2.3 86.5





1.a.4.a other Sectors - Solid Fuels - Commercial/institutional Co2 3 434.8 0.7 91.7


Co2 3 420.3 0.7 92.3

4.D.1 Wastewater treatment and Discharge - Domestic CH4 2 827.7 0.6 92.9

1.a.1.b EnergyIndustries-LiquidFuels-Petroleumrefining Co2 2 802.4 0.6 93.4


Co2 2 497.5 0.5 93.9


N2o 2 330.0 0.5 94.4



Co2 2 114.0 0.4 95.2


N2o 2000.1 0.4 95.6

9.4 Level assessment: 2012 (excluding FolU)

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)



N2o 1791.7 0.4 96.4


HFCs 1396.1 0.3 96.7


N2o 1117.1 0.2 96.9










3.C.2 liming Co2 585.5 0.1 98.5




N2o 489.8 0.1 98.8





N2o 396.9 0.1 99.2





Co2 250.6 0.0 99.4


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)








N2o 141.5 0.0 99.7














CH4 57.9 0.0 99.9








Co2 17.6 0.0 100.0


Co2 17.4 0.0 100.0

9. appendix B


IPCC Category


2000(Gg Co2

eq)

Level Assess-

ment (%)

Cumula-tive total

(%)





3.a.2.f Manure management - Horses CH4 11.6 0.0 100.0



CH4 8.4 0.0 100.0


N2o 8.2 0.0 100.0


2.B Chemical industry - industry S Co2 Ca 0.0 100.0


N2o 4.9 0.0 100.0

1.a.4.b other Sectors - Solid Fuels - residential CH4 4.4 0.0 100.0



1.a.3.a transport - liquid Fuels - Civil aviation CH4 3.3 0.0 100.0



IPCC Cate-gory code

IPCC Category GHG

2000 2012 Trend Assess-ment

(%)

Contri-bution

to trend (%)

Cumu-lative total (%)

(Gg Co2 eq)

3.B.1.bForest land - land Converted to Forest land

Co2 -15339.9 -29490.8 3.9 14.6 14.6


Co2 185027.4 242154.0 3.8 14.1 28.6

1.a.4.b other Sectors - Solid Fuels - residential Co2 3604.2 18461.9 3.2 11.7 40.3


Co2 29101.6 24472.7 2.5 9.1 49.5


Co2 30454.7 29903.3 1.6 6.0 55.4

4.a Solid Waste Disposal CH4 9367.6 18412.0 1.6 5.8 61.3

3.a.1.a enteric Fermentation - Cattle CH4 22819.4 21236.5 1.5 5.4 66.7

2.C.1Metal industry - iron and Steel production

Co2 16410.5 15020.7 1.1 4.1 70.8


N2o 13180.8 11815.0 0.9 3.5 74.3

1.a.4.aother Sectors - liquid Fuels - Commercial/institutional

Co2 7690.4 12685.4 0.7 2.8 77.1


Co2 32623.3 42929.6 0.7 2.7 79.8

2.C.2 Metal industry - Ferroalloys production Co2 8079.1 11623.8 0.4 1.5 81.3

3.B.3.bGrassland - land Converted to Grassland

Co2 7904.1 7871.8 0.4 1.5 82.7


HFCs 0.0 1396.1 0.3 1.2 83.9

1.a.4.bother Sectors - liquid Fuels - residential

Co2 2868.9 2154.6 0.3 1.1 85.0

3.a.1.c enteric Fermentation - Sheep CH4 4169.0 3783.5 0.3 1.1 86.1

3.B.2.b Cropland - land Converted to Cropland Co2 5721.7 5721.0 0.3 1.0 87.1


Co2 1811.2 3434.8 0.3 1.0 88.1

9.5 Trend assessment: 2000 – 2012 (including FolU)

9. appendix B


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)


Co2 3164.8 2802.4 0.2 0.9 89.0

1.a.3.a transport - liquid Fuels - Civil aviation Co2 2040.0 3466.4 0.2 0.8 89.8

3.B.2.aCropland - Cropland remaining Cropland

Co2 -1241.6 -104.9 0.2 0.7 90.6

2.C.3 Metal industry - aluminium production pFCs 982.2 1979.2 0.2 0.7 91.2

2.B Chemical industry - industry F N2o Ca Ca 0.2 0.6 91.8


Co2 2207.2 3420.3 0.2 0.6 92.4


N2o 2479.0 2330.0 0.2 0.6 93.0


Co2 1186.1 2114.0 0.2 0.6 93.6


N2o 1939.5 1791.7 0.1 0.5 94.0

2.B Chemical industry - industry K Co2 Ca Ca 0.1 0.3 94.4

2.B Chemical industry - industry x Co2 Ca Ca 0.1 0.3 94.7


N2o 1934.7 2000.1 0.1 0.3 95.0

3.a.1.d enteric Fermentation - Goats CH4 993.6 855.6 0.1 0.3 95.3

1.a.3.c transport - liquid Fuels - railways Co2 551.5 318.0 0.1 0.3 95.5

3.C.1.cemissions from biomass burning - Grasslands

N2o 637.5 436.2 0.1 0.3 95.8

1.a.4.b other Sectors - Biomass - residential N2o 380.9 146.6 0.1 0.3 96.1

3.C.3 Urea application Co2 211.5 555.0 0.1 0.2 96.3

2.C.3 Metal industry - aluminium production Co2 1105.5 1066.9 0.1 0.2 96.6

3.D.1 Harvested Wood products Co2 -312.2 -511.8 0.1 0.2 96.8

3.B.5.bSettlements - land Converted to Settlements

Co2 1191.0 1183.2 0.1 0.2 97.0

3.B.4.a.Wetlands - Flooded land remaining floodedland

CH4 429.2 267.9 0.1 0.2 97.2

3.a.1.j enteric Fermentation - other CH4 1106.7 1106.7 0.1 0.2 97.4

2.C.6 Metal industry - Zinc production Co2 194.4 0.0 0.1 0.2 97.6


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)

2.a.1 Mineral industry - Cement production Co2 3347.1 3844.5 0.1 0.2 97.8


Co2 171.5 0.0 0.0 0.2 98.0

3.B.6.bother land - land Converted to other land

Co2 -960.2 -960.3 0.0 0.2 98.2


Co2 2217.7 2497.5 0.0 0.2 98.3

1.a.4.b other Sectors - Biomass - residential CH4 222.0 85.5 0.0 0.2 98.5

3.a.2.i Manure management - poultry N2o 642.6 954.6 0.0 0.1 98.6

3.a.2.h Manure management - Swine CH4 612.1 586.8 0.0 0.1 98.8


CH4 542.5 542.3 0.0 0.1 98.9

3.C.2 liming Co2 384.1 585.5 0.0 0.1 99.0

1.a.5.a Non-Specified-LiquidFuels-Stationary Co2 985.6 1110.5 0.0 0.1 99.0


N2o 853.5 1117.1 0.0 0.1 99.1

2.B Chemical industry - industry x CH4 Ca Ca 0.0 0.1 99.2

2.a.2 Mineral industry - lime production Co2 426.4 453.1 0.0 0.1 99.2

1.a.4.b other Sectors - Solid Fuels - residential N2o 16.6 85.2 0.0 0.1 99.3

3.C.1.bemissions from biomass burning - Croplands

CH4 244.2 233.6 0.0 0.1 99.3

3.a.2.a Manure management - Cattle CH4 182.6 167.5 0.0 0.0 99.4


N2o 463.0 618.1 0.0 0.0 99.4

3.C.1.aemissions from biomass burning - Forest lands

N2o 151.1 133.1 0.0 0.0 99.4


N2o 134.2 112.9 0.0 0.0 99.5


N2o 441.3 489.8 0.0 0.0 99.5

1.a.3.c transport - liquid Fuels - railways N2o 63.0 36.3 0.0 0.0 99.6

3.B.1.aForest land - Forest land remaining Forest land

Co2 -6225.2 -4834.4 0.0 0.0 99.6

9. appendix B


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)

3.C.1.eemissions from biomass burning - other lands

CH4 19.7 59.9 0.0 0.0 99.6

2.B Chemical industry - industry Q Co2 Ca Ca 0.0 0.0 99.7


N2o 352.8 396.9 0.0 0.0 99.7


CH4 2348.3 2827.7 0.0 0.0 99.7


CH4 164.6 230.8 0.0 0.0 99.7

2.a.3 Mineral industry - Glass production Co2 74.4 113.9 0.0 0.0 99.8


N2o 81.5 77.9 0.0 0.0 99.8


Co2 188.5 250.6 0.0 0.0 99.8

2.C.5 Metal industry - lead production Co2 39.2 27.3 0.0 0.0 99.8


N2o 23.1 48.1 0.0 0.0 99.8

3.a.2.a Manure management - Cattle N2o 301.0 347.0 0.0 0.0 99.8


CH4 267.6 308.9 0.0 0.0 99.9

3.a.2.h Manure management - Swine N2o 63.7 61.1 0.0 0.0 99.9

3.a.2.i Manure management - poultry CH4 54.2 80.5 0.0 0.0 99.9

3.a.1.h enteric Fermentation - Swine CH4 52.1 50.0 0.0 0.0 99.9


N2o 105.4 141.5 0.0 0.0 99.9

3.C.1.demissions from biomass burning - Wetlands

N2o 21.1 12.5 0.0 0.0 99.9

3.C.6indirect N2o emissions from manure management

N2o 350.0 436.9 0.0 0.0 99.9

3.a.1.f enteric Fermentation - Horses CH4 111.8 127.5 0.0 0.0 99.9

2.D.1Non-energy products from Fuels and SolventUse-Paraffinwaxuse

Co2 7.4 17.4 0.0 0.0 99.9


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)


N2o 18.3 30.2 0.0 0.0 99.9


N2o 572.2 689.1 0.0 0.0 99.9

3.a.1.g enteric Fermentation - Mules and assess CH4 37.7 38.4 0.0 0.0 100.0


9. appendix B


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)

1.a.4.b other Sectors - Solid Fuels - residential Co2 3 604.2 18 461.9 3.4 14.7 14.7


Co2 185 027.4 242 154.0 3.0 13.0 27.7


Co2 29 101.6 24 472.7 2.9 12.3 40.0


Co2 30 454.7 29 903.3 1.9 8.3 48.3

3.a.1.a enteric Fermentation - Cattle CH4 22 819.4 21 236.5 1.7 7.5 55.8

4.a Solid Waste Disposal CH4 9 367.6 18 412.0 1.7 7.1 62.9

2.C.1Metal industry - iron and Steel production

Co2 16 410.5 15 020.7 1.3 5.6 68.5


N2o 13 180.8 11 815.0 1.1 4.8 73.3


Co2 7 690.4 12 685.4 0.8 3.3 76.6


Co2 32 623.3 42 929.6 0.6 2.5 79.2

2.C.2 Metal industry - Ferroalloys production Co2 8 079.1 11 623.8 0.4 1.7 80.8


Co2 2 868.9 2 154.6 0.3 1.5 82.3


HFCs 0.0 1 396.1 0.3 1.5 83.8

3.a.1.c enteric Fermentation – Sheep CH4 4 169.0 3 783.5 0.3 1.5 85.3


Co2 1 811.2 3 434.8 0.3 1.2 86.5


Co2 3 164.8 2 802.4 0.3 1.2 87.7

1.a.3.a transport - liquid Fuels - Civil aviation Co2 2 040.0 3 466.4 0.2 1.0 88.7

2.C.3 Metal industry - aluminium production pFCs 982.2 1 979.2 0.2 0.8 89.5

9.6 Trend assessment: 2000 – 2012 (including FolU)


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)


N2o 2 479.0 2 330.0 0.2 0.8 90.3

2.B Chemical industry - industry F N2o Ca Ca 0.2 0.8 91.0


Co2 2 207.2 3 420.3 0.2 0.7 91.8


Co2 1 186.1 2 114.0 0.2 0.7 92.4


N2o 1 939.5 1 791.7 0.2 0.6 93.1


N2o 1 934.7 2 000.1 0.1 0.4 93.5

2.B Chemical industry - industry x Co2 Ca Ca 0.1 0.4 93.9

2.B Chemical industry - industry K Co2 Ca Ca 0.1 0.4 94.3

3.a.1.d enteric Fermentation - Goats CH4 993.6 855.6 0.1 0.4 94.7

1.a.3.c transport - liquid Fuels - railways Co2 551.5 318.0 0.1 0.4 95.1


N2o 637.5 436.2 0.1 0.4 95.5

1.a.4.b other Sectors - Biomass - residential N2o 380.9 146.6 0.1 0.3 95.8

2.a.1 Mineral industry - Cement production Co2 3347.1 3844.5 0.1 0.3 96.1

2.C.3 Metal industry - aluminium production Co2 1105.5 1066.9 0.1 0.3 96.5

3.C.3 Urea application Co2 211.5 555.0 0.1 0.3 96.8

3.a.1.j enteric Fermentation - other CH4 1106.7 1106.7 0.1 0.3 97.1


Co2 2217.7 2497.5 0.1 0.3 97.3

2.C.6 Metal industry - Zinc production Co2 194.4 0.0 0.1 0.3 97.6


Co2 171.5 0.0 0.1 0.2 97.8

1.a.4.b other Sectors - Biomass - residential CH4 222.0 85.5 0.0 0.2 98.0

3.a.2.h Manure management - Swine CH4 612.1 586.8 0.0 0.2 98.2

3.a.2.i Manure management - poultry N2o 642.6 954.6 0.0 0.2 98.3


CH4 542.5 542.3 0.0 0.1 98.5

9. appendix B


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)

1.a.5.a Non-Specified-LiquidFuels-Stationary Co2 985.6 1110.5 0.0 0.1 98.6

3.C.2 liming Co2 384.1 585.5 0.0 0.1 98.7


CH4 2348.3 2827.7 0.0 0.1 98.8

2.a.2 Mineral industry - lime production Co2 426.4 453.1 0.0 0.1 98.9

2.B Chemical industry - industry x CH4 Ca Ca 0.0 0.1 99.0


CH4 244.2 233.6 0.0 0.1 99.0

1.a.4.b other Sectors - Solid Fuels - residential N2o 16.6 85.2 0.0 0.1 99.1

3.a.2.a Manure management - Cattle CH4 182.6 167.5 0.0 0.1 99.2


N2o 441.3 489.8 0.0 0.1 99.2


N2o 853.5 1117.1 0.0 0.1 99.3


N2o 151.1 133.1 0.0 0.1 99.3


N2o 134.2 112.9 0.0 0.1 99.4


N2o 463.0 618.1 0.0 0.0 99.5

1.a.3.c transport - liquid Fuels - railways N2o 63.0 36.3 0.0 0.0 99.5


N2o 352.8 396.9 0.0 0.0 99.5

2.B Chemical industry - industry Q Co2 Ca Ca 0.0 0.0 99.6


CH4 19.7 59.9 0.0 0.0 99.6

3.a.2.a Manure management - Cattle N2o 301.0 347.0 0.0 0.0 99.6


CH4 164.6 230.8 0.0 0.0 99.7


CH4 267.6 308.9 0.0 0.0 99.7


N2o 81.5 77.9 0.0 0.0 99.7


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)


N2o 572.2 689.1 0.0 0.0 99.7

2.a.3 Mineral industry - Glass production Co2 74.4 113.9 0.0 0.0 99.8

2.C.5 Metal industry - lead production Co2 39.2 27.3 0.0 0.0 99.8


N2o 23.1 48.1 0.0 0.0 99.8

3.a.2.h Manure management - Swine N2o 63.7 61.1 0.0 0.0 99.8


Co2 188.5 250.6 0.0 0.0 99.8

3.a.1.h enteric Fermentation - Swine CH4 52.1 50.0 0.0 0.0 99.9


N2o 21.1 12.5 0.0 0.0 99.9

3.a.2.i Manure management - poultry CH4 54.2 80.5 0.0 0.0 99.9

3.a.1.f enteric Fermentation - Horses CH4 111.8 127.5 0.0 0.0 99.9


N2o 105.4 141.5 0.0 0.0 99.9

3.a.1.g enteric Fermentation - Mules and assess CH4 37.7 38.4 0.0 0.0 99.9

2.D.1Non-energy products from Fuels and SolventUse-Paraffinwaxuse

Co2 7.4 17.4 0.0 0.0 99.9


N2o 18.3 30.2 0.0 0.0 99.9


CH4 18.0 15.6 0.0 0.0 99.9


CH4 11.6 8.4 0.0 0.0 100.0


N2o 8.3 15.8 0.0 0.0 100.0

2.C.2 Metal industry - Ferroalloys production CH4 0.0 3.4 0.0 0.0 100.0

1.a.4.b other Sectors - Solid Fuels - residential CH4 0.9 4.4 0.0 0.0 100.0


N2o 5.2 3.1 0.0 0.0 100.0

9. appendix B


IPCC Cate-gory code

IPCC Category GHG


(%)

Contri-bution

to trend (%)


(Gg Co2 eq)


CH4 44.2 57.9 0.0 0.0 100.0


CH4 7.1 11.7 0.0 0.0 100.0


CH4 7.0 5.8 0.0 0.0 100.0

2.B Chemical industry - industry S Co2 Ca Ca 0.0 0.0 100.0

1.a.3.a transport - liquid Fuels - Civil aviation N2o 5.1 8.6 0.0 0.0 100.0

3.C.6indirect N2o emissions from manure management

N2o 350.0 436.9 0.0 0.0 100.0


N2o 5.3 8.2 0.0 0.0 100.0


N2o 3.2 2.3 0.0 0.0 100.0


Co2 13.0 17.6 0.0 0.0 100.0


N2o 2.8 4.9 0.0 0.0 100.0


CH4 2.2 1.4 0.0 0.0 100.0

3.a.2.f Manure management - Horses CH4 10.2 11.6 0.0 0.0 100.0


N2o 0.8 0.0 0.0 0.0 100.0

1.a.3.a transport - liquid Fuels - Civil aviation CH4 2.0 3.3 0.0 0.0 100.0

3.a.2.g Manure management - Mules and asses CH4 3.4 3.5 0.0 0.0 100.0



10. APPENDIx C: UNCertaiNty aNalySiS

IPCC Category Gas

base year emissions / removals

Year t emissions/ removals

Activity data

uncertainty

EF uncertainty

Combined uncertainty

Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity

Uncertainty in trend in national emissions

introduced by EF/ estimation parameter

uncertainty

Uncertainty in trend in national emissions introduced by activity

data uncertainty

Uncertainty introduced into the

trend in total national emissions

Gg Co2e Gg Co2e (+)% (+)% (+)% (fraction) % % % % %

1.A - Fuel Combustion Activities

1.a.1.a.i - electricity Generation - liquid Fuels

Co2 0.00 2 527.71 5 7 8.6023 0.0026 0.0073 0.0073 0.0513 0.0519 0.0053


CH4 0.00 2.33 5 75 75.1665 0.0000 0.0000 0.0000 0.0005 0.0000 0.0000


N2o 0.00 6.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0013 0.0001 0.0000

1.a.1.a.i - electricity Generation - Solid Fuels

Co2 185 027.44 242 154.03 5 7 8.6023 23.4730 0.0328 0.7025 0.2295 4.9675 24.7291


CH4 44.21 57.87 5 75 75.1665 0.0001 0.0000 0.0002 0.0006 0.0012 0.0000


N2o 853.53 1 117.05 5 75 75.1665 0.0381 0.0002 0.0032 0.0114 0.0229 0.0007

1.A.1.b-PetroleumRefining-liquid Fuels

Co2 3 164.79 2 802.41 5 7 8.6023 0.0031 0.0033 0.0081 0.0233 0.0575 0.0038


CH4 1.67 1.35 5 75 75.1665 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000


N2o 3.15 2.32 5 75 75.1665 0.0000 0.0000 0.0000 0.0004 0.0000 0.0000

1.a.1.c.i - Manufacture of Solid Fuels - liquid Fuels

Co2 30 454.73 29 903.32 5 7 8.6023 0.3580 0.0234 0.0868 0.1640 0.6134 0.4032


CH4 11.64 8.42 5 75 75.1665 0.0000 0.0000 0.0000 0.0013 0.0002 0.0000


N2o 105.38 141.49 5 75 75.1665 0.0006 0.0000 0.0004 0.0022 0.0029 0.0000

1.a.1.c.i - Manufacture of Solid Fuels - Solid Fuels

Co2 0.00 0.00 5 7 8.6023 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10.1 Energy sector uncertainty

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




1.A - Fuel Combustion Activities


Co2 0.00 2 527.71 5 7 8.6023 0.0026 0.0073 0.0073 0.0513 0.0519 0.0053


CH4 0.00 2.33 5 75 75.1665 0.0000 0.0000 0.0000 0.0005 0.0000 0.0000


N2o 0.00 6.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0013 0.0001 0.0000


Co2 185 027.44 242 154.03 5 7 8.6023 23.4730 0.0328 0.7025 0.2295 4.9675 24.7291


CH4 44.21 57.87 5 75 75.1665 0.0001 0.0000 0.0002 0.0006 0.0012 0.0000


N2o 853.53 1 117.05 5 75 75.1665 0.0381 0.0002 0.0032 0.0114 0.0229 0.0007


Co2 3 164.79 2 802.41 5 7 8.6023 0.0031 0.0033 0.0081 0.0233 0.0575 0.0038


CH4 1.67 1.35 5 75 75.1665 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000


N2o 3.15 2.32 5 75 75.1665 0.0000 0.0000 0.0000 0.0004 0.0000 0.0000


Co2 30 454.73 29 903.32 5 7 8.6023 0.3580 0.0234 0.0868 0.1640 0.6134 0.4032


CH4 11.64 8.42 5 75 75.1665 0.0000 0.0000 0.0000 0.0013 0.0002 0.0000


N2o 105.38 141.49 5 75 75.1665 0.0006 0.0000 0.0004 0.0022 0.0029 0.0000


Co2 0.00 0.00 5 7 8.6023 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





N2o 0.00 0.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.1.c.ii - other energy industries - liquid Fuels

Co2 0.00 0.00 10 7 12.2066 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.00 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.2 - Manufacturing industries and Construction - liquid Fuels

Co2 1 186.10 2 114.02 10 7 12.2066 0.0036 0.0018 0.0061 0.0129 0.0867 0.0077


CH41.07 1.90 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 2.75 4.90 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0002 0.0000

1.a.2 - Manufacturing industries and Construction - Solid Fuels

Co2 29 101.57 24 472.72 10 7 12.2066 0.4827 0.0343 0.0710 0.2400 1.0041 1.0657


CH4 6.95 5.85 10 75 75.6637 0.0000 0.0000 0.0000 0.0006 0.0002 0.0000


N2o 134.25 112.89 10 75 75.6637 0.0004 0.0002 0.0003 0.0119 0.0046 0.0002

1.a.2 - Manufacturing industries and Construction - Gaseous Fuels

Co2 2 217.75 2 497.46 10 7 12.2066 0.0050 0.0008 0.0072 0.0055 0.1025 0.0105


CH4 0.91 1.02 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 1.17 1.32 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0001 0.0000

1.a.3.a.i - international aviation (international Bunkers) - liquid Fuels

Co2 2 972.40 2 413.48 5 1.5 5.2202 0.0009 0.0038 0.0070 0.0056 0.0495 0.0025

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





N2o 0.00 0.00 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 0.00 0.00 10 7 12.2066 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.00 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 1 186.10 2 114.02 10 7 12.2066 0.0036 0.0018 0.0061 0.0129 0.0867 0.0077


CH41.07 1.90 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 2.75 4.90 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0002 0.0000


Co2 29 101.57 24 472.72 10 7 12.2066 0.4827 0.0343 0.0710 0.2400 1.0041 1.0657


CH4 6.95 5.85 10 75 75.6637 0.0000 0.0000 0.0000 0.0006 0.0002 0.0000


N2o 134.25 112.89 10 75 75.6637 0.0004 0.0002 0.0003 0.0119 0.0046 0.0002


Co2 2 217.75 2 497.46 10 7 12.2066 0.0050 0.0008 0.0072 0.0055 0.1025 0.0105


CH4 0.91 1.02 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 1.17 1.32 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0001 0.0000


Co2 2 972.40 2 413.48 5 1.5 5.2202 0.0009 0.0038 0.0070 0.0056 0.0495 0.0025


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





CH4 2.87 2.33 5 50 50.2494 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000


N2o 7.38 5.99 5 50 50.2494 0.0000 0.0000 0.0000 0.0005 0.0001 0.0000

1.a.3.a.ii - Domestic aviation - liquid Fuels

Co2 2 040.00 3 466.36 5 1.5 5.2202 0.0018 0.0027 0.0101 0.0040 0.0711 0.0051


CH4 1.97 3.35 5 50 50.2494 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 5.07 8.61 5 50 50.2494 0.0000 0.0000 0.0000 0.0003 0.0002 0.0000

1.a.3.b - road transportation - liquid Fuels

Co232 623.34 42 929.58 5 2 5.3852 0.2891 0.0065 0.1245 0.0130 0.8807 0.7757


CH4 267.59 308.90 5 9 10.2956 0.0001 0.0001 0.0009 0.0006 0.0063 0.0000


N2o 463.05 618.12 5 72 72.1734 0.0108 0.0001 0.0018 0.0085 0.0127 0.0002

1.a.3.c - railways - liquid Fuels Co2 551.45 317.96 5 2 5.3852 0.0000 0.0011 0.0009 0.0021 0.0065 0.0000

1.a.3.c - railways - liquid Fuels CH4 0.71 0.41 5 9 10.2956 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.3.c - railways - liquid Fuels N2o 63.00 36.33 5 72 72.1734 0.0000 0.0001 0.0001 0.0088 0.0007 0.0001

1.a.4.a - Commercial/institutional - liquid Fuels

Co2 7 690.43 12 685.42 10 7 12.2066 0.1297 0.0090 0.0368 0.0628 0.5205 0.2748


CH4 7.10 11.74 10 75 75.6637 0.0000 0.0000 0.0000 0.0006 0.0005 0.0000


N2o 18.27 30.22 10 75 75.6637 0.0000 0.0000 0.0001 0.0016 0.0012 0.0000

1.a.4.a - Commercial/institutional - Solid Fuels

Co2 1 811.19 3 434.78 10 7 12.2066 0.0095 0.0034 0.0100 0.0239 0.1409 0.0204


CH4 0.44 0.82 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0000 0.0000


N2o 8.31 15.84 10 75 75.6637 0.0000 0.0000 0.0000 0.0012 0.0007 0.0000

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





CH4 2.87 2.33 5 50 50.2494 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000


N2o 7.38 5.99 5 50 50.2494 0.0000 0.0000 0.0000 0.0005 0.0001 0.0000


Co2 2 040.00 3 466.36 5 1.5 5.2202 0.0018 0.0027 0.0101 0.0040 0.0711 0.0051


CH4 1.97 3.35 5 50 50.2494 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 5.07 8.61 5 50 50.2494 0.0000 0.0000 0.0000 0.0003 0.0002 0.0000


Co232 623.34 42 929.58 5 2 5.3852 0.2891 0.0065 0.1245 0.0130 0.8807 0.7757


CH4 267.59 308.90 5 9 10.2956 0.0001 0.0001 0.0009 0.0006 0.0063 0.0000


N2o 463.05 618.12 5 72 72.1734 0.0108 0.0001 0.0018 0.0085 0.0127 0.0002

1.a.3.c - railways - liquid Fuels Co2 551.45 317.96 5 2 5.3852 0.0000 0.0011 0.0009 0.0021 0.0065 0.0000

1.a.3.c - railways - liquid Fuels CH4 0.71 0.41 5 9 10.2956 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.3.c - railways - liquid Fuels N2o 63.00 36.33 5 72 72.1734 0.0000 0.0001 0.0001 0.0088 0.0007 0.0001


Co2 7 690.43 12 685.42 10 7 12.2066 0.1297 0.0090 0.0368 0.0628 0.5205 0.2748


CH4 7.10 11.74 10 75 75.6637 0.0000 0.0000 0.0000 0.0006 0.0005 0.0000


N2o 18.27 30.22 10 75 75.6637 0.0000 0.0000 0.0001 0.0016 0.0012 0.0000


Co2 1 811.19 3 434.78 10 7 12.2066 0.0095 0.0034 0.0100 0.0239 0.1409 0.0204


CH4 0.44 0.82 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0000 0.0000


N2o 8.31 15.84 10 75 75.6637 0.0000 0.0000 0.0000 0.0012 0.0007 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




1.a.4.a - Commercial/institutional - Gaseous Fuels

Co2 12.96 17.62 10 7 12.2066 0.0000 0.0000 0.0001 0.0000 0.0007 0.0000


CH4 0.01 0.01 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.01 0.01 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.b - residential - liquid Fuels

Co2 2 868.85 2 154.59 10 7 12.2066 0.0037 0.0041 0.0063 0.0289 0.0884 0.0087


CH4 2.19 1.41 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0001 0.0000


N2o 5.18 3.11 10 75 75.6637 0.0000 0.0000 0.0000 0.0007 0.0001 0.0000

1.a.4.b - residential - Solid Fuels Co2 3 604.18 18 461.91 10 7 12.2066 0.2747 0.0405 0.0536 0.2836 0.7575 0.6542

1.a.4.b - residential - Solid Fuels CH4 0.86 4.41 10 75 75.6637 0.0000 0.0000 0.0000 0.0007 0.0002 0.0000

1.a.4.b - residential - Solid Fuels N2o 16.63 85.16 10 75 75.6637 0.0002 0.0002 0.0002 0.0140 0.0035 0.0002

1.a.4.b - residential - Biomass Co2 0.00 0.00 40 7 40.6079 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.b - residential - Biomass CH4 221.96 85.46 40 75 85.0000 0.0003 0.0006 0.0002 0.0416 0.0140 0.0019

1.a.4.b - residential - Biomass N2o 380.87 146.64 40 75 85.0000 0.0008 0.0010 0.0004 0.0715 0.0241 0.0057

1.a.4.c.i - Stationary - liquid Fuels

Co2 2 207.17 3 420.32 10 7 12.2066 0.0094 0.0019 0.0099 0.0135 0.1403 0.0199


CH4 2.07 3.20 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 5.32 8.23 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0003 0.0000

1.a.4.c.i - Stationary - Solid Fuels Co2 171.52 0.00 10 7 12.2066 0.0000 0.0006 0.0000 0.0043 0.0000 0.0000

1.a.4.c.i - Stationary - Solid Fuels CH4 0.04 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.c.i - Stationary - Solid Fuels N2o 0.79 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000

1.a.4.c.ii - off-road vehicles and other Machinery - liquid Fuels

Co2 0.00 0.00 2 2 2.8284 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 9 9.2195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





Co2 12.96 17.62 10 7 12.2066 0.0000 0.0000 0.0001 0.0000 0.0007 0.0000


CH4 0.01 0.01 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.01 0.01 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 2 868.85 2 154.59 10 7 12.2066 0.0037 0.0041 0.0063 0.0289 0.0884 0.0087


CH4 2.19 1.41 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0001 0.0000


N2o 5.18 3.11 10 75 75.6637 0.0000 0.0000 0.0000 0.0007 0.0001 0.0000

1.a.4.b - residential - Solid Fuels Co2 3 604.18 18 461.91 10 7 12.2066 0.2747 0.0405 0.0536 0.2836 0.7575 0.6542

1.a.4.b - residential - Solid Fuels CH4 0.86 4.41 10 75 75.6637 0.0000 0.0000 0.0000 0.0007 0.0002 0.0000

1.a.4.b - residential - Solid Fuels N2o 16.63 85.16 10 75 75.6637 0.0002 0.0002 0.0002 0.0140 0.0035 0.0002

1.a.4.b - residential - Biomass Co2 0.00 0.00 40 7 40.6079 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.b - residential - Biomass CH4 221.96 85.46 40 75 85.0000 0.0003 0.0006 0.0002 0.0416 0.0140 0.0019

1.a.4.b - residential - Biomass N2o 380.87 146.64 40 75 85.0000 0.0008 0.0010 0.0004 0.0715 0.0241 0.0057


Co2 2 207.17 3 420.32 10 7 12.2066 0.0094 0.0019 0.0099 0.0135 0.1403 0.0199


CH4 2.07 3.20 10 75 75.6637 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000


N2o 5.32 8.23 10 75 75.6637 0.0000 0.0000 0.0000 0.0003 0.0003 0.0000

1.a.4.c.i - Stationary - Solid Fuels Co2 171.52 0.00 10 7 12.2066 0.0000 0.0006 0.0000 0.0043 0.0000 0.0000

1.a.4.c.i - Stationary - Solid Fuels CH4 0.04 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.c.i - Stationary - Solid Fuels N2o 0.79 0.00 10 75 75.6637 0.0000 0.0000 0.0000 0.0002 0.0000 0.0000


Co2 0.00 0.00 2 2 2.8284 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 9 9.2195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





N2o 0.00 0.00 2 72 72.0278 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.4.c.iii - Fishing (mobile combustion) - liquid Fuels

Co2 0.00 0.00 2 2 2.8284 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 9 9.2195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.00 0.00 2 72 72.0278 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.5.a - Stationary - liquid Fuels Co2 985.58 1 110.53 5 7 8.6023 0.0005 0.0003 0.0032 0.0024 0.0228 0.0005

1.a.5.a - Stationary - liquid Fuels CH40.98 1.11 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.5.a - Stationary - liquid Fuels N2o 2.53 2.85 5 75 75.1665 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000

1.a.3.b.vi - Urea-based catalysts Co2 0.00 0.00 30 7 30.8058 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.b.1 - Fugitive Emissions from Fuels - Solid Fuels

1.B.1.a.i.1 - Mining Co2 19.18 16.74 10 63 63.7887 0.0000 0.0000 0.0000 0.0013 0.0007 0.0000

1.B.1.a.i.1 - Mining CH4 1 603.94 1 409.65 10 63 63.7887 0.0437 0.0017 0.0041 0.1080 0.0578 0.0150

1.B.1.a.i.2 - post-mining seam gas emissions

Co2 4.48 3.91 10 50 50.9902 0.0000 0.0000 0.0000 0.0002 0.0002 0.0000


CH4 374.95 329.53 10 50 50.9902 0.0015 0.0004 0.0010 0.0200 0.0135 0.0006

1.B.1.a.i.3 - abandoned underground mines

CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.i.4 - Flaring of drained methane or conversion of methane to Co2

CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.ii.1 - Mining Co2 0.00 0.00 10 63 63.7887 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.ii.1 - Mining CH4 0.00 0.00 10 63 63.7887 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.ii.2 - post-mining seam gas emissions

Co2 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





N2o 0.00 0.00 2 72 72.0278 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 0.00 0.00 2 2 2.8284 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 9 9.2195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


N2o 0.00 0.00 2 72 72.0278 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.5.a - Stationary - liquid Fuels Co2 985.58 1 110.53 5 7 8.6023 0.0005 0.0003 0.0032 0.0024 0.0228 0.0005

1.a.5.a - Stationary - liquid Fuels CH40.98 1.11 5 75 75.1665 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.a.5.a - Stationary - liquid Fuels N2o 2.53 2.85 5 75 75.1665 0.0000 0.0000 0.0000 0.0001 0.0001 0.0000

1.a.3.b.vi - Urea-based catalysts Co2 0.00 0.00 30 7 30.8058 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.b.1 - Fugitive Emissions from Fuels - Solid Fuels

1.B.1.a.i.1 - Mining Co2 19.18 16.74 10 63 63.7887 0.0000 0.0000 0.0000 0.0013 0.0007 0.0000

1.B.1.a.i.1 - Mining CH4 1 603.94 1 409.65 10 63 63.7887 0.0437 0.0017 0.0041 0.1080 0.0578 0.0150


Co2 4.48 3.91 10 50 50.9902 0.0000 0.0000 0.0000 0.0002 0.0002 0.0000


CH4 374.95 329.53 10 50 50.9902 0.0015 0.0004 0.0010 0.0200 0.0135 0.0006

1.B.1.a.i.3 - abandoned underground mines

CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.i.4 - Flaring of drained methane or conversion of methane to Co2

CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.ii.1 - Mining Co2 0.00 0.00 10 63 63.7887 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.1.a.ii.1 - Mining CH4 0.00 0.00 10 63 63.7887 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 10 50 50.9902 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




1.b.2 - Fugitive Emissions from Fuels - Oil and Natural Gas

1.B.2.a.i - venting Co2 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.i - venting CH4 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.i - venting N2o 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.ii - Flaring Co2 752.04 641.83 25 75 79.0569 0.0139 0.0009 0.0019 0.0645 0.0658 0.0085

1.B.2.a.ii - Flaring CH4 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.3 - other emissions from energy production

Co2 28 146.58 25 136.44 25 75 79.0569 21.3619 0.0289 0.0729 2.1680 2.5782 11.3476


CH4 2 450.60 2 684.17 25 75 79.0569 0.2436 0.0011 0.0078 0.0811 0.2753 0.0824

1.C - CO2 Transport Injection and Storage

TOTAL

Sum 344 695.0945 429 955.4182 46.76 39.45

Uncertain-ty in total inventory

6.84 Trend uncertainty 6.28

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




1.b.2 - Fugitive Emissions from Fuels - Oil and Natural Gas

1.B.2.a.i - venting Co2 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.i - venting CH4 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.i - venting N2o 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.B.2.a.ii - Flaring Co2 752.04 641.83 25 75 79.0569 0.0139 0.0009 0.0019 0.0645 0.0658 0.0085

1.B.2.a.ii - Flaring CH4 0.00 0.00 25 75 79.0569 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 28 146.58 25 136.44 25 75 79.0569 21.3619 0.0289 0.0729 2.1680 2.5782 11.3476


CH4 2 450.60 2 684.17 25 75 79.0569 0.2436 0.0011 0.0078 0.0811 0.2753 0.0824

1.C - CO2 Transport Injection and Storage

TOTAL

Sum 344 695.0945 429 955.4182 46.76 39.45




IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




2.A - Mineral Industry

2.a.1 - Cement production Co2 3 347.05 3 844.50 1 4.5 4.6098 0.0001 4.4815 0.1140 20.1670 0.1612 406.7328

2.a.2 - lime production Co2 426.37 453.09 1 6 6.0828 0.0000 0.5725 0.0134 3.4349 0.0190 11.7986

2.a.3 - Glass production Co2 74.36 113.91 5 60 60.2080 0.0000 0.0988 0.0034 5.9289 0.0239 35.1527

2.a.4.a - Ceramics Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.b - other Uses of Soda ash Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.c - Non Metallurgical Mag-nesia production

Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.d - other (please specify) Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.b - Chemical Industry

2.B.1 - ammonia production Co2 485.28 217.80 5 6 7.8102 0.0000 0.6604 0.0065 3.9623 0.0457 15.7020

2.B.1 - ammonia production CH4 71.81 70.10 5 6 7.8102 0.0000 0.0966 0.0021 0.5796 0.0147 0.3362

2.B.2 - Nitric acid production N2o 1 570.25 745.00 2 10 10.1980 0.0000 2.1350 0.0221 21.3498 0.0625 455.8180

2.B.3 - adipic acid production N2o 0.00 0.00 2 10 10.1980 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.4 - Caprolactam, Glyoxal and Glyoxylic acid production

N2o 0.00 0.00 2 40 40.0500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.5 - Carbide production Co2 1.98 5.20 5 10 11.1803 0.0000 0.0026 0.0002 0.0257 0.0011 0.0007

2.B.5 - Carbide production CH4 0.00 0.00 5 10 11.1803 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.6 - titanium Dioxide production

Co2 437.62 158.66 5 10 11.1803 0.0000 0.5967 0.0047 5.9665 0.0333 35.6007

2.B.7 - Soda ash production Co2 0.00 0.00 5 10 11.1803 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.a - Methanol Co2 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.a - Methanol CH4 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.b - ethylene Co2 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.b - ethylene CH4 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.c - ethylene Dichloride and vinyl Chloride Monomer

Co2 0.00 0.00 2 20 20.0998 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 10 10.1980 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.d - ethylene oxide Co2 0.00 0.00 10 10 14.1421 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10.2 IPPU sector uncertainty

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




2.A - Mineral Industry

2.a.1 - Cement production Co2 3 347.05 3 844.50 1 4.5 4.6098 0.0001 4.4815 0.1140 20.1670 0.1612 406.7328

2.a.2 - lime production Co2 426.37 453.09 1 6 6.0828 0.0000 0.5725 0.0134 3.4349 0.0190 11.7986

2.a.3 - Glass production Co2 74.36 113.91 5 60 60.2080 0.0000 0.0988 0.0034 5.9289 0.0239 35.1527

2.a.4.a - Ceramics Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.b - other Uses of Soda ash Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.c - Non Metallurgical Mag-nesia production

Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.a.4.d - other (please specify) Co2 0.00 0.00 43 2.5 43.0726 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.b - Chemical Industry

2.B.1 - ammonia production Co2 485.28 217.80 5 6 7.8102 0.0000 0.6604 0.0065 3.9623 0.0457 15.7020

2.B.1 - ammonia production CH4 71.81 70.10 5 6 7.8102 0.0000 0.0966 0.0021 0.5796 0.0147 0.3362

2.B.2 - Nitric acid production N2o 1 570.25 745.00 2 10 10.1980 0.0000 2.1350 0.0221 21.3498 0.0625 455.8180

2.B.3 - adipic acid production N2o 0.00 0.00 2 10 10.1980 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.4 - Caprolactam, Glyoxal and Glyoxylic acid production

N2o 0.00 0.00 2 40 40.0500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.5 - Carbide production Co2 1.98 5.20 5 10 11.1803 0.0000 0.0026 0.0002 0.0257 0.0011 0.0007

2.B.5 - Carbide production CH4 0.00 0.00 5 10 11.1803 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.6 - titanium Dioxide production

Co2 437.62 158.66 5 10 11.1803 0.0000 0.5967 0.0047 5.9665 0.0333 35.6007

2.B.7 - Soda ash production Co2 0.00 0.00 5 10 11.1803 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.a - Methanol Co2 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.a - Methanol CH4 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.b - ethylene Co2 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.b - ethylene CH4 0.00 0.00 30 30 42.4264 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


Co2 0.00 0.00 2 20 20.0998 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CH4 0.00 0.00 2 10 10.1980 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.d - ethylene oxide Co2 0.00 0.00 10 10 14.1421 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




2.B.8.d - ethylene oxide CH4 0.00 0.00 10 60 60.8276 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.e - acrylonitrile Co2 0.00 0.00 10 10 14.1421 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.e - acrylonitrile CH4 0.00 0.00 10 60 60.8276 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.f - Carbon Black Co2 138.57 134.88 10 15 18.0278 0.0000 0.1864 0.0040 2.7966 0.0566 7.8243

2.B.8.f - Carbon Black CH4 0.07 0.07 10 85 85.5862 0.0000 0.0001 0.0000 0.0083 0.0000 0.0001

2.D - Non-Energy Products from Fuels and Solvent Use

2.C.1 - iron and Steel production Co2 16 410.53 15 020.70 5 25 25.4951 0.0601 22.0012 0.4454 550.0305 3.1498 302 543.4944

2.C.1 - iron and Steel production CH4 0.00 0.00 5 25 25.4951 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.2 - Ferroalloys production Co2 8 079.14 11 623.80 5 25 25.4951 0.0360 10.7331 0.3447 268.3278 2.4375 72 005.7599

2.C.2 - Ferroalloys production CH4 3.62 3.37 5 25 25.4951 0.0000 0.0049 0.0001 0.1217 0.0007 0.0148

2.C.3 - aluminium production Co2 1 105.47 1 066.90 5 10 11.1803 0.0001 1.4872 0.0316 14.8717 0.2237 221.2178

2.C.3 - aluminium production CF4 982.17 1 970.91 5 190 190.0658 0.0575 1.2910 0.0584 245.2944 0.4133 60 169.5366

2.C.3 - aluminium production C2F6 156.94 325.74 5 190 190.0658 0.0016 0.2060 0.0097 39.1434 0.0683 1 532.2117

2.C.4 - Magnesium production Co2 0.00 0.00 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.4 - Magnesium production SF6 0.00 0.00 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.5 - lead production Co2 39.16 27.29 5 15 15.8114 0.0000 0.0530 0.0008 0.7951 0.0057 0.6321

2.C.6 - Zinc production Co2 194.36 0.00 10 50 50.9902 0.0000 0.2671 0.0000 13.3551 0.0000 178.3593

2.C - Metal Industry

2.D.1 - lubricant Use Co2 188.48 250.58 10 50 50.9902 0.0001 0.2516 0.0074 12.5797 0.1051 158.2591

2.D.2-ParaffinWaxUse Co2 7.44 17.42 10 100 100.4988 0.0000 0.0097 0.0005 0.9703 0.0073 0.9415

2.E - Electronics Industry

2.F - Product Uses as Substitutes for Ozone Depleting Substances

2.F.1. - refrigeration and Stationary air Conditioning

HFC-23 0.00 7 400.00 25 25 35.3553 0.0280 0.2194 0.2194 5.4862 7.7587 90.2968


HFC-32 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-41 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-43-10mee

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty




2.B.8.d - ethylene oxide CH4 0.00 0.00 10 60 60.8276 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.e - acrylonitrile Co2 0.00 0.00 10 10 14.1421 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.e - acrylonitrile CH4 0.00 0.00 10 60 60.8276 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.B.8.f - Carbon Black Co2 138.57 134.88 10 15 18.0278 0.0000 0.1864 0.0040 2.7966 0.0566 7.8243

2.B.8.f - Carbon Black CH4 0.07 0.07 10 85 85.5862 0.0000 0.0001 0.0000 0.0083 0.0000 0.0001

2.D - Non-Energy Products from Fuels and Solvent Use

2.C.1 - iron and Steel production Co2 16 410.53 15 020.70 5 25 25.4951 0.0601 22.0012 0.4454 550.0305 3.1498 302 543.4944

2.C.1 - iron and Steel production CH4 0.00 0.00 5 25 25.4951 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.2 - Ferroalloys production Co2 8 079.14 11 623.80 5 25 25.4951 0.0360 10.7331 0.3447 268.3278 2.4375 72 005.7599

2.C.2 - Ferroalloys production CH4 3.62 3.37 5 25 25.4951 0.0000 0.0049 0.0001 0.1217 0.0007 0.0148

2.C.3 - aluminium production Co2 1 105.47 1 066.90 5 10 11.1803 0.0001 1.4872 0.0316 14.8717 0.2237 221.2178

2.C.3 - aluminium production CF4 982.17 1 970.91 5 190 190.0658 0.0575 1.2910 0.0584 245.2944 0.4133 60 169.5366

2.C.3 - aluminium production C2F6 156.94 325.74 5 190 190.0658 0.0016 0.2060 0.0097 39.1434 0.0683 1 532.2117

2.C.4 - Magnesium production Co2 0.00 0.00 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.4 - Magnesium production SF6 0.00 0.00 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.C.5 - lead production Co2 39.16 27.29 5 15 15.8114 0.0000 0.0530 0.0008 0.7951 0.0057 0.6321

2.C.6 - Zinc production Co2 194.36 0.00 10 50 50.9902 0.0000 0.2671 0.0000 13.3551 0.0000 178.3593

2.C - Metal Industry

2.D.1 - lubricant Use Co2 188.48 250.58 10 50 50.9902 0.0001 0.2516 0.0074 12.5797 0.1051 158.2591

2.D.2-ParaffinWaxUse Co2 7.44 17.42 10 100 100.4988 0.0000 0.0097 0.0005 0.9703 0.0073 0.9415

2.E - Electronics Industry

2.F - Product Uses as Substitutes for Ozone Depleting Substances


HFC-23 0.00 7 400.00 25 25 35.3553 0.0280 0.2194 0.2194 5.4862 7.7587 90.2968


HFC-32 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-41 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-43-10mee

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





HFC-125 0.00 30 450.00 25 25 35.3553 0.4746 0.9030 0.9030 22.5752 31.9261 1 528.9155


HFC-134 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-134a

0.00 1 355 640.00 25 25 35.3553 940.6369 40.2020 40.2020 1 005.0512 1 421.3570 3 030 383.7487


HFC-152a

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-143 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-143a

0.00 133 206.00 25 25 35.3553 9.0820 3.9503 3.9503 98.7569 139.6634 29 258.7961


HFC-227ea

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236fa

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-245ca

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-152 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-161 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236cb

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236ea

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-245fa

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-365mfc

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CF4 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





HFC-125 0.00 30 450.00 25 25 35.3553 0.4746 0.9030 0.9030 22.5752 31.9261 1 528.9155


HFC-134 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-134a

0.00 1 355 640.00 25 25 35.3553 940.6369 40.2020 40.2020 1 005.0512 1 421.3570 3 030 383.7487


HFC-152a

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-143 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-143a

0.00 133 206.00 25 25 35.3553 9.0820 3.9503 3.9503 98.7569 139.6634 29 258.7961


HFC-227ea

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236fa

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-245ca

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-152 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-161 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236cb

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-236ea

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-245fa

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


HFC-365mfc

0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


CF4 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





C2F6 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C3F8 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C4F10 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


c-C4F8 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C5F12 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C6F14 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.G - Electrical Equipment

2.H - Other

TOTAL

Sum 33 720.67011 562

745.9346950.38 3 499 041.15


30.83 Trend uncertainty 1 870.57

Excluding 2.F - Product Uses as Substitutes for Ozone Depleting Substances



10. appendix C


IPCC Category Gas



Activity data

uncertainty

EF uncertainty


Contri-bution to

variance in Year t

Type A sensitivity

Type b sensitivity



uncertainty


data uncertainty





C2F6 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C3F8 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C4F10 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


c-C4F8 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C5F12 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


C6F14 0.00 0.00 25 25 35.3553 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2.G - Electrical Equipment

2.H - Other

TOTAL

Sum 33 720.67011 562

745.9346950.38 3 499 041.15


30.83 Trend uncertainty 1 870.57

Excluding 2.F - Product Uses as Substitutes for Ozone Depleting Substances




11. APPENDIx D: eNerGy SeCtor – CoMpariSoN oF tHe reFereNCe vS SeCtoral approaCH

REFERENCE VS SECTORAL APPROACH

PeriodReference approach Sectoral Approach Difference Difference

MtCo2 Mt %

2000 368 305 62.87 21%

2001 361 304 57.23 19%

2002 351 315 35.68 11%

2003 402 339 63.14 19%

2004 430 355 75.76 21%

2005 422 351 70.66 20%

2006 433 358 74.36 21%

2007 463 386 76.43 20%

2008 500 379 121.54 32%

2009 490 387 103.34 27%

2010 483 401 82.59 21%

2011 477 382 95.69 25%

2012 471 394 77.72 20%

Average 76.69 21%

11. appendix D


the reference approach is a top-down approach, using a country’s energy supply data to calculate the emissions of Co2 from combustion of mainly fossil fuels. the reference approach was applied on the basis of relatively easily available energy supply statistics. it is good practice to apply both a sectoral approach and the reference approach to estimate a country’s Co2 emissions from fuel combustion and to compare the resultsofthesetwoindependentestimates.Significantdifferences may indicate possible problems with the activitydata,netcalorificvalues,carboncontent,excluded carbon calculation etc.

the reference approach and the Sectoral approach often have different results because the reference approach is a top-down approach using a country’s energy supply data and has no detailed information on how the individual fuels are used in each sector. the average difference in the reference and sectoral approach is approximately 21%whichisquitesignificant.Themainreasonsforthedifferences were mainly from the gaseous fuels and solid fuels which had an average difference of 124% and 43% respectively. the generic reasons for the difference in reference and sectoral approach are as follows:

• Missing information on stock changes that may occuratthefinalconsumerlevel.Therelevanceofconsumer stocks depends on the method used for the Sectoral approach.

• High distribution losses for gas will cause the reference approach to be higher than the Sectoral approach,

• Unrecorded consumption of gas or other fuels may lead to an underestimation of the Sectoral approach.

• the treatment of transfers and reclassifications of energy products may cause a difference in the Sectoral approach estimation since different net

calorificvaluesandemissionfactorsmaybeuseddependingonhowthefuelisclassified.

• NetCalorificValues(NCV)used inthesectoralapproach differs from those used in the reference approach. in power generation, NCv values in the sectoral approach vary over the 2000-2010 time series based on the information provided by industry;

• activity data on liquid fuels in the sectoral approach particularly for energy industries is sourced directly from the companies involved and has been reconciled with other publicly available datasets;

• allocation of solid fuels between energy use, non-energy use as well use for synfuels production remains one of the key drives for the differences observed between the two datasets.

• inconsistencies on the sources of activity data within the time series and in some cases the application of extrapolation

• the misallocation of the quantities of fuels used for conversion into derived products (other than power or heat) or quantities combusted in the energy sector.

• SimplificationsintheReferenceApproach.Therearesmall quantities of carbon which should be included in the reference approach because their emissions fall under fuel combustion. these quantities have beenexcludedwheretheflowsaresmallornotrepresented by a major statistic available within energy data.

• examples of quantities not accounted for in the reference approach include lubricants used in two-stroke e.g. natural gas leakage from pipelines, emissions from energy transformation, etc. likely to be <5% of the reference approach.


Solid Fuels

SOLID FUELS


tJ tJ %

2000 3 386 284 2 282 887 1 103 396.84 48%

2001 3 380 674 2 258 656 1 122 017.53 50%

2002 3 314 815 2 363 553 951 262.16 40%

2003 3 687 775 2 576 732 1 111 043.31 43%

2004 3 886 355 2 702 932 1 183 423.23 44%

2005 3 849 218 2 689 024 1 160 194.32 43%

2006 3 918 677 2 744 792 1 173 885.33 43%

2007 4 039 592 2 978 118 1 061 474.11 36%

2008 4 417 995 2 918 928 1 499 067.10 51%

2009 4 503 950 2 984 812 1 519 138.10 51%

2010 4 227 203 3 085 055 1 142 147.62 37%

2011 4 122 114 2 878 821 1 243 292.94 43%

2012 4 063 666 2 997 646 1 066 020.21 36%

Average 1 179 720.22 43%

11. appendix D


the source of discrepancies on solid fuels between the reference and sectoral approach can mainly be attributed to three categories namely; electricity generation, manufacturing industries, construction and residential sector.

a. Main electricity and heat production: Data on fuel consumption for public electricity generation was obtained directly from the national power utility for the period 2000 to 2012 for the sectoral approach. Within this sector is the inclusion of auto electricity producers which was sourced from the energy balance for the period 2000-2006 and extrapolated for the period 2007-2012. Discrepancies in solid fuels can be as a result of the extrapolation for auto-electricity producers.

b. Manufacturing industries and construction: Sources of activity data in the manufacturing and construction sector were sourced from the energy digest from the Doe for solid fuels for the period 2000-2007. For

remaining period of 2007-2010 the SaMi report was used to extrapolate the fuel consumption from this category. the difference in the two approaches may have been caused by the use of difference sources of activity data within the time series.

c. Residential sector: solid fuel consumption data from the residential sector was sourced from a variety of sources such as Doe energy digest reports, DMr South african Mineral industry, Food and agriculture organization and Sapia. the Doe energy reports were used to source solid fuels for the period 2000 to 2006, for the remaining period (20007-2010) the SaMi report was used to extrapolate the consumption of solid fuels for this category. Discrepancies in the solid fuels consumed in the residential sector might have been caused by the inconsistencies in the data sources within the time series.


Solid Fuels

LIQUID FUELS


tJ tJ %

2000 601 099 752 782 151 682.97 20%

2001 505 640 760 331 254 690.98 33%

2002 451 170 766 941 315 771.88 41%

2003 690 236 801 598 111 362.42 14%

2004 759 904 855 739 95 834.86 11%

2005 692 790 855 010 162 220.33 19%

2006 755 250 881 793 126 543.33 14%

2007 967 528 934 230 33 297.61 3.6%

2008 989 048 917 143 71 904.76 7.8%

2009 767 022 933 975 166 952.95 18%

2010 1 001 273 995 940 5 332.86 0.5%

2011 1 059 595 1 019 469 40 125.98 3.9%

2012 1 048 417 1 028 408 20 008.23 1.9%

Average 119 671.48 15%

11. appendix D


the source of discrepancies on liquid fuels between the reference and sectoral approach can mainly be attributed to three categories namely; road transportation, commercial sector and agriculture/Forestry/Fishing/Fish Farms.

a. Themainconsumingsectorofliquidfuelsistheroadtransportation. the main source of activity data for liquid fuel consumption in the transportation sector, was the Sapia annual reports. the discrepancies between the sectoral and reference approach might be attributed to the fact that Sapia reports liquid fuel consumption at a disaggregated level various users.

b. Commercial sector: the commercial sector refers to institutional or commercial building. the main source of activity data for liquid fuels for this category was the Sapia for the period 2000-2006. an extrapolation method was applied to estimate activity data for the period 2007-2012, this might have also contributed to the differences in the reference and sectoral approach.

c. Agriculture/Forestry/Fishing/Fish Farms: Data on fuel consumptionintheagriculture,forestry,fishingandfishfarmswasobtainedfromtheEnergydigestbythe Doe (mostly solid fuels) and Sapia for liquid fuels. Discrepancies might have been attributed by the application of different sources of activity data for the time series.


LIQUID FUELS


tJ tJ %

2000 58 490 39 763 18 726.60 47%

2001 75 990 41 482 34 508.42 83%

2002 75 362 43 299 32 063.40 74%

2003 45 176 48 844 -3 668.43 7.5%

2004 116 393 50 671 65 722.04 130%

2005 115 472 53 541 61 930.94 116%

2006 112 039 56 433 55 605.77 99%

2007 160 103 59 323 100 780.23 170%

2008 163 578 62 213 101 365.28 163%

2009 128 702 65 101 63 601.23 98%

2010 162 155 68 888 93 266.76 135%

2011 160 982 51 744 109 238.46 211%

2012 168 979 44 832 124 147.25 277%

Average 65 945.23 124%

Gaseous Fuels

11. appendix D


the source of discrepancies on gaseous fuels between the reference and sectoral approach can mainly be attributed to the manufacturing industries and construction mainly in the production of synfuels.

GHG National inventory report for South africa 2000 -2012 331


12. APPENDIx E: aGriCUltUral aCtivity Data

Livestock category

Livestock sub-category from Du Toit et al. (2013b) Livestock sub-category in this inventory

Sheep

Non-wool breeding eweNon-wool ewe

Non-wool young ewe

Non-wool breeding ramNon-wool ram

Non-wool young ram

Non-wool weanersNon-wool lamb

Non-wool lamb

Merino breeding ewe

Wool ewe

Merino young ewe

other wool breeding ewe

other wool young ewe

Karakul breeding ewe

Karakul young ewe

Merino breeding ram

Wool ram

Merino young ram

other wool breeding ram

other wool young ram

Karakul breeding ram

Karakul young ram

Merino weaner

Wool lamb

Merino lamb

other wool weaner

other wool lamb

Karakul weaner

Karakul lamb

Goats

Breeding buckBuck

young buck

Breeding doeDoe

young doe

12.1 Livestock sub-categories for sheep, goats and swine

12. appendix e


Livestock category

Livestock sub-category from Du Toit et al. (2013b) Livestock sub-category in this inventory

Goats

WeanerKid

Kid

Breeding buck

angora

young buck

Breeding doe

young doe

Weaner

Kid

Breeding buck

Milk goat

young buck

Breeding doe

young doe

Weaner

Kid

Swine

Boars

Boarsreplacement boars

Cull boars

Dry gestating sow

Sowslactating sow

replacement sow

Cull sow

pre-wean piglets piglets

Baconers Baconers

porkers porkers


Category Sub-categoryPopulation

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Dairy (Commercial - tMr)

lactating cow 437 637 437 637 375 749 349 226 340 385 362 487 353 646 349 226 433 217 442 058 442 058

Dry cow 106 452 106 452 91 398 84 947 82 796 88 173 86 022 84 947 105 377 107 528 107 528

lactating heifer 42 092 40 985 39 877 31 015 27 692 31 015 31 015 32 123 35 446 37 662 37 662

pregnant heifer 56 663 55 172 53 680 41 751 37 278 41 751 41 751 43 243 47 716 50 698 50 698

Heifer >1yr 19 427 18 916 18 405 14 315 12 781 14 315 14 315 14 826 16 360 17 382 17 382

Heifer 6-12mths 38 854 37 832 36 809 28 630 25 562 28 630 28 630 29 652 32 720 34 764 34 764

Heifer 2-6mths 25 903 25 221 24 540 19 086 17 041 19 086 19 086 19 768 21 813 23 176 23 176

Calves <6mths 25 903 25 221 24 540 19 086 17 041 19 086 19 086 19 768 21 813 23 176 23 176

Total TMR 752 932 747 436 664 998 588 056 560 577 604 544 593 552 593 552 714 461 736 445 736 445

Dairy (Commercial - pasture)

lactating cow 358 667 358 667 307 946 286 209 278 963 297 078 289 832 286 209 355 044 362 290 362 290

Dry cow 87 243 87 243 74 906 69 618 67 856 72 262 70 500 69 618 86 362 88 125 88 125


pregnant heifer 46 438 45 216 43 994 34 218 30 551 34 218 34 218 35 440 39 106 41 550 41 550

Heifer >1yr 15 922 15 503 15 084 11 732 10 475 11 732 11 732 12 151 13 408 14 246 14 246

Heifer 6-12mths 31 843 31 005 30 167 23 463 20 950 23 463 23 463 24 301 26 815 28 491 28 491

Heifer 2-6mths 21 229 20 670 20 112 15 642 13 966 15 642 15 642 16 201 17 877 18 994 18 994

Calves <6mths 21 229 20 670 20 112 15 642 13 966 15 642 15 642 16 201 17 877 18 994 18 994

Total Pasture 617 068 612 564 545 002 481 944 459 423 495 456 486 448 486 448 585 539 603 555 603 555

Total Dairy 1 370 000 1 360 000 1 210 000 1 070 000 1 020 000 1 100 000 1 080 000 1 080 000 1 300 000 1 340 000 1 340 000

other cattle (Commercial)

Feedlot 420 000 420 000 420 000 420 000 420 000 420 000 391 148 400 819 399 822 461 800 484 274

Bulls 200 000 190 000 190 000 190 000 200 000 190 000 180 000 150 000 160 000 180 000 200 000

Cows 3 540 000 3 600 000 3 430 000 3 080 000 2 840 000 3 140 000 3 080 000 2 460 000 2 710 000 2 390 000 2 980 000

Heifers 1 050 000 1 040 000 1 020 000 950 000 1 390 000 920 000 900 000 850 000 770 000 700 000 910 000

ox 230 000 240 000 200 000 260 000 280 000 210 000 200 000 170 000 240 000 280 000 170 000

young ox 660 000 660 000 540 000 570 000 520 000 510 000 500 000 1 080 000 1 140 000 860 000 630 000

Calves 1 630 000 1 610 000 1 470 000 1 910 000 1 770 000 2 110 000 2 070 000 2 400 000 1 960 000 2 490 000 1 990 000

Total commercial 7 730 000 7 760 000 7 270 000 7 380 000 7 420 000 7 500 000 7 350 000 7 530 000 7 371 148 7 300 819 7 279 822

12.2 Livestock population data

12. appendix e



2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Dairy (Commercial - tMr)

lactating cow 437 637 437 637 375 749 349 226 340 385 362 487 353 646 349 226 433 217 442 058 442 058

Dry cow 106 452 106 452 91 398 84 947 82 796 88 173 86 022 84 947 105 377 107 528 107 528


pregnant heifer 56 663 55 172 53 680 41 751 37 278 41 751 41 751 43 243 47 716 50 698 50 698

Heifer >1yr 19 427 18 916 18 405 14 315 12 781 14 315 14 315 14 826 16 360 17 382 17 382

Heifer 6-12mths 38 854 37 832 36 809 28 630 25 562 28 630 28 630 29 652 32 720 34 764 34 764

Heifer 2-6mths 25 903 25 221 24 540 19 086 17 041 19 086 19 086 19 768 21 813 23 176 23 176

Calves <6mths 25 903 25 221 24 540 19 086 17 041 19 086 19 086 19 768 21 813 23 176 23 176

Total TMR 752 932 747 436 664 998 588 056 560 577 604 544 593 552 593 552 714 461 736 445 736 445

Dairy (Commercial - pasture)

lactating cow 358 667 358 667 307 946 286 209 278 963 297 078 289 832 286 209 355 044 362 290 362 290

Dry cow 87 243 87 243 74 906 69 618 67 856 72 262 70 500 69 618 86 362 88 125 88 125


pregnant heifer 46 438 45 216 43 994 34 218 30 551 34 218 34 218 35 440 39 106 41 550 41 550

Heifer >1yr 15 922 15 503 15 084 11 732 10 475 11 732 11 732 12 151 13 408 14 246 14 246

Heifer 6-12mths 31 843 31 005 30 167 23 463 20 950 23 463 23 463 24 301 26 815 28 491 28 491

Heifer 2-6mths 21 229 20 670 20 112 15 642 13 966 15 642 15 642 16 201 17 877 18 994 18 994

Calves <6mths 21 229 20 670 20 112 15 642 13 966 15 642 15 642 16 201 17 877 18 994 18 994

Total Pasture 617 068 612 564 545 002 481 944 459 423 495 456 486 448 486 448 585 539 603 555 603 555

Total Dairy 1 370 000 1 360 000 1 210 000 1 070 000 1 020 000 1 100 000 1 080 000 1 080 000 1 300 000 1 340 000 1 340 000

other cattle (Commercial)

Feedlot 420 000 420 000 420 000 420 000 420 000 420 000 391 148 400 819 399 822 461 800 484 274

Bulls 200 000 190 000 190 000 190 000 200 000 190 000 180 000 150 000 160 000 180 000 200 000

Cows 3 540 000 3 600 000 3 430 000 3 080 000 2 840 000 3 140 000 3 080 000 2 460 000 2 710 000 2 390 000 2 980 000

Heifers 1 050 000 1 040 000 1 020 000 950 000 1 390 000 920 000 900 000 850 000 770 000 700 000 910 000

ox 230 000 240 000 200 000 260 000 280 000 210 000 200 000 170 000 240 000 280 000 170 000

young ox 660 000 660 000 540 000 570 000 520 000 510 000 500 000 1 080 000 1 140 000 860 000 630 000

Calves 1 630 000 1 610 000 1 470 000 1 910 000 1 770 000 2 110 000 2 070 000 2 400 000 1 960 000 2 490 000 1 990 000




2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

other cattle (Communal)

Bulls 137 359 130 491 130 491 130 491 137 359 130 491 123 623 103 019 109 887 123 623 137 359

Cows 2 431 256 2 472 464 2 355 709 2 115 330 1 950 499 2 156 538 2 115 330 1 689 517 1 861 216 1 641 441 2 046 651

Heifers 721 135 714 267 700 531 652 456 954 646 631 852 618 116 583 776 528 833 480 757 624 984

ox 157 963 164 831 137 359 178 567 192 303 144 227 137 359 116 755 164 831 192 303 116 755

young ox 453 285 453 285 370 870 391 473 357 134 350 266 343 398 741 739 782 947 590 644 432 681

Calves 1 119 477 1 105 741 1 009 589 1 311 779 1 215 628 1 449 138 1 421 667 1 648 309 1 346 119 1 710 121 1 366 723

Total communal 5 020 475 5 041 079 4 704 549 4 780 097 4 807 568 4 862 512 4 759 493 4 883 116 4 793 833 4 738 889 4 725 153

Total Other cattle 12 750 475 12 801 079 11 974 549 12 160 096 12 227 569 12 362 512 12 109 493 12 413 115 12 164 981 12 039 708 12 004 975

Sheep (Commercial)

Wool lamb 6 610 800 6 446 000 6 427 600 6 476 000 6 395 200 6 407 600 6 219 600 6 299 200 6 389 200 6 296 000 6 174 400

Wool ewe 9 420 390 9 185 550 9 159 330 9 228 300 9 113 160 9 130 830 8 862 930 8 976 360 9 104 610 8 971 800 8 798 520

Wool ram 495 810 483 450 482 070 485 700 479 640 480 570 466 470 472 440 479 190 472 200 463 080

Non-wool lamb 2 823 600 2 753 200 2 618 000 2 601 200 2 520 400 2 486 800 2 558 400 2 470 400 2 408 800 2 470 800 2 422 800

Non-wool ewe 4 023 630 3 923 310 3 730 650 3 706 710 3 591 570 3 543 690 3 645 720 3 520 320 3 432 540 3 520 890 3 452 490

Non-wool ram 211 770 206 490 196 350 195 090 189 030 186 510 191 880 185 280 180 660 185 310 181 710


Sheep (Communal)

Wool lamb 922 602 899 602 897 034 903 789 892 513 894 243 868 006 879 115 891 675 878 668 861 698

Wool ewe 1 314 708 1 281 933 1 278 274 1 287 900 1 271 831 1 274 297 1 236 909 1 252 739 1 270 637 1 252 102 1 227 920

Wool ram 69 195 67 470 67 278 67 784 66 938 67 068 65 100 65 934 66 876 65 900 64 627

Non-wool lamb 394 061 384 236 365 368 363 023 351 746 347 057 357 050 344 768 336 172 344 824 338 125

Non-wool ewe 561 537 547 536 520 649 517 308 501 239 494 557 508 796 491 295 479 045 491 375 481 829

Non-wool ram 29 555 28 818 27 403 27 227 26 381 26 029 26 779 25 858 25 213 25 862 25 359

Total communal 3 291 657 3 209 596 3 156 005 3 167 030 3 110 648 3 103 251 3 062 639 3 059 709 3 069 617 3 058 732 2 999 558

Total sheep 26 877 657 26 207 596 25 770 005 25 860 030 25 399 648 25 339 251 25 007 639 24 983 709 25 064 617 24 975 732 24 492 558

Goats (Commercial)

Buck 207 086 213 417 194 863 189 939 190 291 187 828 191 786 186 070 185 894 182 640 180 442

Doe 654 760 674 778 616 114 600 544 601 657 593 872 606 383 588 311 587 755 577 468 570 517

Kids 618 749 637 667 582 229 567 515 568 566 561 210 573 033 555 955 555 429 545 708 539 140

angora 856 557 882 745 806 001 785 632 787 087 776 903 793 270 769 629 768 901 755 444 746 351

Milk goats 17 847 18 392 16 793 16 369 16 399 16 187 16 528 16 035 16 020 15 740 15 550


12. appendix e



2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

other cattle (Communal)

Bulls 137 359 130 491 130 491 130 491 137 359 130 491 123 623 103 019 109 887 123 623 137 359

Cows 2 431 256 2 472 464 2 355 709 2 115 330 1 950 499 2 156 538 2 115 330 1 689 517 1 861 216 1 641 441 2 046 651

Heifers 721 135 714 267 700 531 652 456 954 646 631 852 618 116 583 776 528 833 480 757 624 984

ox 157 963 164 831 137 359 178 567 192 303 144 227 137 359 116 755 164 831 192 303 116 755

young ox 453 285 453 285 370 870 391 473 357 134 350 266 343 398 741 739 782 947 590 644 432 681

Calves 1 119 477 1 105 741 1 009 589 1 311 779 1 215 628 1 449 138 1 421 667 1 648 309 1 346 119 1 710 121 1 366 723

Total communal 5 020 475 5 041 079 4 704 549 4 780 097 4 807 568 4 862 512 4 759 493 4 883 116 4 793 833 4 738 889 4 725 153

Total Other cattle 12 750 475 12 801 079 11 974 549 12 160 096 12 227 569 12 362 512 12 109 493 12 413 115 12 164 981 12 039 708 12 004 975

Sheep (Commercial)

Wool lamb 6 610 800 6 446 000 6 427 600 6 476 000 6 395 200 6 407 600 6 219 600 6 299 200 6 389 200 6 296 000 6 174 400

Wool ewe 9 420 390 9 185 550 9 159 330 9 228 300 9 113 160 9 130 830 8 862 930 8 976 360 9 104 610 8 971 800 8 798 520

Wool ram 495 810 483 450 482 070 485 700 479 640 480 570 466 470 472 440 479 190 472 200 463 080

Non-wool lamb 2 823 600 2 753 200 2 618 000 2 601 200 2 520 400 2 486 800 2 558 400 2 470 400 2 408 800 2 470 800 2 422 800

Non-wool ewe 4 023 630 3 923 310 3 730 650 3 706 710 3 591 570 3 543 690 3 645 720 3 520 320 3 432 540 3 520 890 3 452 490

Non-wool ram 211 770 206 490 196 350 195 090 189 030 186 510 191 880 185 280 180 660 185 310 181 710


Sheep (Communal)

Wool lamb 922 602 899 602 897 034 903 789 892 513 894 243 868 006 879 115 891 675 878 668 861 698

Wool ewe 1 314 708 1 281 933 1 278 274 1 287 900 1 271 831 1 274 297 1 236 909 1 252 739 1 270 637 1 252 102 1 227 920

Wool ram 69 195 67 470 67 278 67 784 66 938 67 068 65 100 65 934 66 876 65 900 64 627

Non-wool lamb 394 061 384 236 365 368 363 023 351 746 347 057 357 050 344 768 336 172 344 824 338 125

Non-wool ewe 561 537 547 536 520 649 517 308 501 239 494 557 508 796 491 295 479 045 491 375 481 829

Non-wool ram 29 555 28 818 27 403 27 227 26 381 26 029 26 779 25 858 25 213 25 862 25 359

Total communal 3 291 657 3 209 596 3 156 005 3 167 030 3 110 648 3 103 251 3 062 639 3 059 709 3 069 617 3 058 732 2 999 558

Total sheep 26 877 657 26 207 596 25 770 005 25 860 030 25 399 648 25 339 251 25 007 639 24 983 709 25 064 617 24 975 732 24 492 558

Goats (Commercial)

Buck 207 086 213 417 194 863 189 939 190 291 187 828 191 786 186 070 185 894 182 640 180 442

Doe 654 760 674 778 616 114 600 544 601 657 593 872 606 383 588 311 587 755 577 468 570 517

Kids 618 749 637 667 582 229 567 515 568 566 561 210 573 033 555 955 555 429 545 708 539 140

angora 856 557 882 745 806 001 785 632 787 087 776 903 793 270 769 629 768 901 755 444 746 351

Milk goats 17 847 18 392 16 793 16 369 16 399 16 187 16 528 16 035 16 020 15 740 15 550




2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Goats (Communal)

Buck 147 130 151 628 138 445 134 947 135 197 133 447 136 259 132 198 132 073 129 761 128 200

Doe 2 795 507 2 880 974 2 630 507 2 564 032 2 568 780 2 535 542 2 588 960 2 511 801 2 509 427 2 465 506 2 435 830

Kids 1 709 028 1 761 278 1 608 155 1 567 516 1 570 419 1 550 099 1 582 756 1 535 585 1 534 134 1 507 283 1 489 140

Total communal 4 651 664 4 793 880 4 377 107 4 266 494 4 274 395 4 219 089 4 307 974 4 179 584 4 175 634 4 102 550 4 053 170

Total goats 7 006 664 7 220 880 6 593 107 6 426 494 6 438 395 6 355 089 6 488 974 6 295 584 6 289 634 6 179 550 6 105 170

Swine (commercial)

Boars 30 563 31 138 31 732 30 860 30 860 30 637 30 099 30 637 29 969 29 932 29 579

Sows 529 757 539 728 550 021 534 903 534 903 531 043 521 715 531 043 519 464 518 821 512 709

piglets 543 340 553 567 564 124 548 619 548 619 544 660 535 093 544 660 532 784 532 124 525 856

porkers 81 501 83 035 84 619 82 293 82 293 81 699 80 264 81 699 79 918 79 819 78 878

Baconers 461 839 470 532 479 505 466 326 466 326 462 961 454 829 462 961 452 866 452 305 446 977


Swine (communal)

Boars 7 868 8 017 8 169 7 945 7 945 7 888 7 749 7 888 7 716 7 706 7 615

Sows 102 290 104 216 106 203 103 284 103 284 102 539 100 738 102 539 100 303 100 179 98 999

piglets 52 457 53 444 54 463 52 966 52 966 52 584 51 660 52 584 51 437 51 374 50 769

porkers 7 868 8 017 8 169 7 945 7 945 7 888 7 749 7 888 7 716 7 706 7 615

Baconers 44 588 45 427 46 294 45 021 45 021 44 696 43 911 44 696 43 722 43 668 43 153

Total communal 215 072 219 120 223 299 217 161 217 161 215 594 211 807 215 594 210 893 210 632 208 151

Total swine 1 862 072 1 897 120 1 933 299 1 880 161 1 880 161 1 866 594 1 833 807 1 866 594 1 825 893 1 823 632 1 802 151

Horses Total 270 000 270 000 270 000 270 000 270 000 270 000 280 000 290 000 298 000 300 000 300 000

Mules and asses asses 150 000 150 000 150 000 150 000 150 000 150 000 150 000 150 500 150 500 150 500 152 000

Mules 14 000 14 000 14 000 14 000 14 000 14 000 14 050 14 100 14 200 14 300 14 300

Totalmules/asses 164 000 164 000 164 000 164 000 164 000 164 000 164 050 164 600 164 700 164 800 166 300

poultry (Commercial) annual avg broilers 65 716 784 65 716 888 67 671 233 71 163 068 72 846 641 79 846 641 86 769 951 89 769 951 90 950 137 92 129 907 93 057 627

layers 17 400 000 17 800 000 17 700 000 17 000 000 17 590 000 18 660 000 20 590 000 22 780 000 23 080 000 22 230 000 23 100 000


poultry (Communal) annual avg broilers 2 760 105 2 760 109 2 842 192 2 988 849 3 050 695 3 353 559 3 625 500 3 770 338 3 819 906 3 869 456 3 908 420

Total poultry 85 876 889 86 276 997 88 213 425 91 151 917 93 276 290 101 860 200 110 536 925 116 320 289 117 850 043 118 229 363 120 066047

Total livestock 136 177 756 136 197 670 136 128 386 138 982 699 140 676 063 149 317 645 157 500 889 163 413 893 164 957 871 165 052 786 166 277 200

12. appendix e



2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Goats (Communal)

Buck 147 130 151 628 138 445 134 947 135 197 133 447 136 259 132 198 132 073 129 761 128 200

Doe 2 795 507 2 880 974 2 630 507 2 564 032 2 568 780 2 535 542 2 588 960 2 511 801 2 509 427 2 465 506 2 435 830

Kids 1 709 028 1 761 278 1 608 155 1 567 516 1 570 419 1 550 099 1 582 756 1 535 585 1 534 134 1 507 283 1 489 140

Total communal 4 651 664 4 793 880 4 377 107 4 266 494 4 274 395 4 219 089 4 307 974 4 179 584 4 175 634 4 102 550 4 053 170

Total goats 7 006 664 7 220 880 6 593 107 6 426 494 6 438 395 6 355 089 6 488 974 6 295 584 6 289 634 6 179 550 6 105 170

Swine (commercial)

Boars 30 563 31 138 31 732 30 860 30 860 30 637 30 099 30 637 29 969 29 932 29 579

Sows 529 757 539 728 550 021 534 903 534 903 531 043 521 715 531 043 519 464 518 821 512 709

piglets 543 340 553 567 564 124 548 619 548 619 544 660 535 093 544 660 532 784 532 124 525 856

porkers 81 501 83 035 84 619 82 293 82 293 81 699 80 264 81 699 79 918 79 819 78 878

Baconers 461 839 470 532 479 505 466 326 466 326 462 961 454 829 462 961 452 866 452 305 446 977


Swine (communal)

Boars 7 868 8 017 8 169 7 945 7 945 7 888 7 749 7 888 7 716 7 706 7 615

Sows 102 290 104 216 106 203 103 284 103 284 102 539 100 738 102 539 100 303 100 179 98 999

piglets 52 457 53 444 54 463 52 966 52 966 52 584 51 660 52 584 51 437 51 374 50 769

porkers 7 868 8 017 8 169 7 945 7 945 7 888 7 749 7 888 7 716 7 706 7 615

Baconers 44 588 45 427 46 294 45 021 45 021 44 696 43 911 44 696 43 722 43 668 43 153

Total communal 215 072 219 120 223 299 217 161 217 161 215 594 211 807 215 594 210 893 210 632 208 151

Total swine 1 862 072 1 897 120 1 933 299 1 880 161 1 880 161 1 866 594 1 833 807 1 866 594 1 825 893 1 823 632 1 802 151

Horses Total 270 000 270 000 270 000 270 000 270 000 270 000 280 000 290 000 298 000 300 000 300 000

Mules and asses asses 150 000 150 000 150 000 150 000 150 000 150 000 150 000 150 500 150 500 150 500 152 000

Mules 14 000 14 000 14 000 14 000 14 000 14 000 14 050 14 100 14 200 14 300 14 300

Totalmules/asses 164 000 164 000 164 000 164 000 164 000 164 000 164 050 164 600 164 700 164 800 166 300

poultry (Commercial) annual avg broilers 65 716 784 65 716 888 67 671 233 71 163 068 72 846 641 79 846 641 86 769 951 89 769 951 90 950 137 92 129 907 93 057 627

layers 17 400 000 17 800 000 17 700 000 17 000 000 17 590 000 18 660 000 20 590 000 22 780 000 23 080 000 22 230 000 23 100 000


poultry (Communal) annual avg broilers 2 760 105 2 760 109 2 842 192 2 988 849 3 050 695 3 353 559 3 625 500 3 770 338 3 819 906 3 869 456 3 908 420

Total poultry 85 876 889 86 276 997 88 213 425 91 151 917 93 276 290 101 860 200 110 536 925 116 320 289 117 850 043 118 229 363 120 066047

Total livestock 136 177 756 136 197 670 136 128 386 138 982 699 140 676 063 149 317 645 157 500 889 163 413 893 164 957 871 165 052 786 166 277 200


13. APPENDIx F: laND SeCtor aCtivity Data

DescriptionClimate Type Area (ha)

WtD WtM CtD CtM total (ha)

indigenous Forest 365 000 11 100 100 0 376 200

thicket /Dense bush 7 460 300 24 800 1 800 400 7 487 300

Woodland/open bush 11 422 900 14 600 3 400 200 11 441 100

low shrubland 38 197 100 60 300 6 500 7 300 38 271 200

plantations / Woodlots 1 707 500 26 900 0 0 1 734 400

Cultivated commercial annual crops pivot 720 900 0 0 0 720 900

Cultivated commercial annual crops non-pivot 9 811 600 500 0 0 9 812 100

Cultivated commercial permanent orchards 312 200 2 200 0 0 314 400

Cultivated commercial permanent vines 174 800 400 0 0 175 200

Cultivated subsistence crops 1 875 600 700 0 0 1 876 300

Settlements 2 663 400 1 000 0 0 2 664 400

Wetlands 936 700 2 100 1 100 300 940 200

Grasslands 23 610 500 63 400 68 100 19 500 23 761 500

Mines 302 000 100 0 0 302 100

Waterbodies (incl coastal sea extent) 447 900 700 0 0 448 600

Bare Ground 11 753 000 3 300 1 800 2 600 11 760 700

Degraded 868 400 100 0 0 868 500

13.1 Land cover by climate

the 2013-14 land cover map (Geoterraimage, 2015) was intersected with the long term climate map provided by the arC.

13. appendix F


Description

Area on each IPCC soil type (ha)

Sandy mineral

soils

Wetland mineral

soils

Spodic mineral

soils

rocky areas

low activity clay

mineral soils

High activity clay

mineral soils

Water body

indigenous Forest 96 696 226 0 30 403 91 295 203 150 1 863

thicket /Dense bush 658 937 924 2 200 1 185 382 562 013 5 817 654 9 370

Woodland/open bush 2 135 567 161 518 1 424 284 576 444 8 243 359 4 073

low shrubland 15 748 119 124 2 063 4 634 043 24 304 20 955 223 7 588

plantations / Woodlots 164 824 36 52 23 253 1 121 099 561 068 568

Cultivated commercial annual crops pivot

162 477 540 0 1 117 50 719 565 961 71

Cultivated commercial annual crops non-pivot

1 167 966 1 252 1 501 57 126 712 547 8 669 075 658

Cultivated commercial permanent orchards

75 632 0 22 6 955 57 575 206 523 242

Cultivated commercial permanent vines

55 978 0 0 4 672 0 128 053 8

Cultivated subsistence crops

110 808 720 0 11 934 175 470 1 737 450 169

Settlements 349 109 539 0 55 124 300 620 2 201 493 394

Wetlands 62 972 1 861 118 28 006 149 993 766 409 9 770

Grasslands 1 644 147 3 601 50 1 724 952 2 210 455 20 142 593 15 568

Mines 78 653 265 0 8 713 43 353 195 486 7

Waterbodies (incl coastal sea extent)

29 812 983 1 9 446 23 205 192 189 199 991

Bare Ground 2 038 559 11 754 12 1 198 308 15 371 9 485 819 3 235

Degraded 166 972 10 0 23 355 15 931 737 663 9

13.2 Land cover by soil type

The2013-14landcovermap(GeoTerraImage,2015)wasintersectedwiththesoiltype(usingtheIPCCdefaultclassification)map provided by the arC.


13.3 biomass and soil data

total above ground biomass data from the NtCSa (Department of environmental affairs, 2015) for the various land classes.

above ground woody biomass data for the more detailed vegetation classes determined by intersecting the NtCSa (Department of environmental affairs, 2015) carbon layers with the 2013-14 land-cover data set.

DescriptionMean biomass

carbon [gC/m^2]Std. Dev. [gC/m^2] tC/ha tdm/ha

indigenous Forest 7186 3 423 71.9 152.9

thicket 2 370 3 159 23.7 50.4

Savanna 418 756 4.2 8.9

plantations 4 603 969 46.0 97.9

Cultivated land 186 50 1.9 3.9

Grasslands 532 748 5.3 11.3

Description

Woody biomass

Mean Woody Biomass [gC/m^2]

Std. Dev. [gC/m^2]

tC/ha tdm/ha

indigenous Forest 828 2 697 8.28 17.62

thicket /Dense bush 3 175 4 697 31.75 67.55

Woodland/open bush 686 1 852 6.86 14.59

low shrubland 248 1 054 2.48 5.29

plantations / Woodlots 8 962 8 826 89.62 190.68

Cultivated commercial annual crops pivot 450 1 122 4.50 9.58

Cultivated commercial annual crops non-pivot 461 1 450 4.61 9.81

Cultivated commercial permanent orchards 2 729 4 591 27.29 58.06

Cultivated commercial permanent vines 1 012 1 354 10.12 21.53

Cultivated subsistence crops 853 2 139 8.53 18.15

Settlements 1 177 2 763 11.77 25.04

Wetlands 981 2 488 9.81 20.87

Grasslands 709 1 873 7.09 15.09

Mines 241 934 2.41 5.13

Waterbodies (incl coastal sea extent) 854 2 555 8.54 18.18

Bare Ground 50 560 0.50 1.07

Degraded 211 565 2.11 4.48

13. appendix F


Herbaceous biomass data for the more detailed vegetation classes determined by intersection the NtCSa (Department of environmental affairs, 2015) layers with the 2013-14 land-cover data set.

Description

Herbaceous biomass

Mean Herbaceous Biomass C [gC/

m^2]

Std. Dev. [gC/m^2]

tC/ha tdm/ha

indigenous Forest 87 82 0.87 1.86

thicket /Dense bush 39 44 0.39 0.82

Woodland/open bush 13 19 0.13 0.27

low shrubland 3 10 0.03 0.06

plantations / Woodlots 81 76 0.81 1.73

Cultivated commercial annual crops pivot 10 15 0.10 0.21

Cultivated commercial annual crops non-pivot 9 17 0.09 0.20

Cultivated commercial permanent orchards 37 46 0.37 0.78

Cultivated commercial permanent vines 12 17 0.12 0.26

Cultivated subsistence crops 15 23 0.15 0.33

Settlements 22 30 0.22 0.46

Wetlands 17 28 0.17 0.37

Grasslands 12 21 0.12 0.27

Mines 8 13 0.08 0.17

Waterbodies (incl coastal sea extent) 18 30 0.18 0.38

Bare Ground 1 6 0.01 0.02

Degraded 6 8 0.06 0.13


Soil carbon data for the various vegetation classes determined by intersection the NtCSa (Department of environmental affairs, 2015) layers with the 2013-14 land-cover data set.

Description

Soil carbon (0 – 30cm)

Mean Soil Carbon Content [gC/m^2]

Std. Dev. [gC/m^2] tC/ha

indigenous Forest 3 184 4 169 31.84

thicket /Dense bush 2 589 3 700 25.89

Woodland/open bush 1 723 2 874 17.23

low shrubland 826 2 542 8.26

plantations / Woodlots 3 795 3 543 37.95

Cultivated commercial annual crops pivot 1 923 3 409 19.23

Cultivated commercial annual crops non-pivot 2 038 3 363 20.38

Cultivated commercial permanent orchards 2 222 3 322 22.22

Cultivated commercial permanent vines 1 117 1 575 11.17

Cultivated subsistence crops 2 614 3 696 26.14

Settlements 2 594 3 680 25.94

Wetlands 2 556 4 300 25.56

Grasslands 2 514 4 326 25.14

Mines 1 543 3 346 15.43

Waterbodies (incl coastal sea extent) 1 296 4 241 12.96

Bare Ground 380 1 574 3.80

Degraded 1 488 2 250 14.88

13. appendix F


14. APPENDIx G: MetHoDoloGy For laND Cover aND laND USe CHaNGe Matrix

14.1 South African 2013-14 National Land-cover dataset

2013 - 2014 South African National Land-Cover Dataset

Dea/CarDNo SCpF002: implementation of land-Use Maps for South africa

Project Specific Data User Report and Metadata

Data proDUCt CreateD ByGEOTERRAIMAGE (South Africa),

www.geoterraimage.comJuly 2015, version 05b (pivot code corrected data)

OPEN ACCESS LICENCE


LICENSE AND COPYRIGHT

the data supplied are licensed by GeoterraiMaGe to the Client, but remains the property of GeoterraiMaGe as is protected by copyright laws. the data can be used by the client as per the conditions set forth in the offer. the client does not obtain any ownership or intellectual property rights in the data. “Client” refers to the South african Department of environmental affairs.

the following shall always be clearly mentioned: (© GeoterraiMaGe - 2014).

OPEN DATA LICENSING (PERMITTED & INTENDED DATA USE)

the Department is granted an open Data User license, which allows the Department unrestricted access to and use thesupplieddata.Thedataisdefinedasthe72ClassGTISouthAfricanNationalLandCoverDataset(2013/2014).Theopen Data license allows the Department to distribute the data unrestricted to third parties with the full understanding that the Department and third parties may not use the data to develop new products that will compete directly with GeoterraiMaGe existing or “in-progress” commercial data products; and that the Department and third parties accept and understand the stated Data Copyright, licensing and liability conditions.

LIAbILITY

the Client has to verify prior to his commitment, the adequacy of GeoterraiMaGe datasets and products to the specificdataandanalysesneedsassetoutbytheclient.GEOTERRAIMAGEwillinnoeventbeliableforanydirect,indirect,incidentalorconsequentialdamagesincludinglostsavingsorprofit,legalfees,lostdata,lostsales,lossofcommitments, business interruption, including loss of or other damage, resulting from its datasets and products, whether or not GeoterraiMaGe has been advised of the possibility of such damage. this includes damages incurred by the client or any third party.

GeoterraiMaGe datasets and products are processed from sources like topographic maps and satellite images. the quality of these sources is not the responsibility of GeoterraiMaGe. the data and data products are provided “as is”, without warranty of any kind, either express or implied, including, but not limited to, the implied warranties of merchantabilityandfitnessforaparticularpurpose.

Whilst all possible care and attention has been taken in the production of the data, neither GeoterraiMaGe or any of its respective sub-organisations or employees, accept any liability whatsoever for any perceived or actual inaccuracies or misrepresentations in the supplied data or accompanying report(s), as a result of the nature of mapping from coarse and medium scale satellite imagery. any mapped boundaries either implied or inferred within the data do not have any legal status whatsoever.

Allrightsnotspecificallygrantedorspecifiedinthisagreement,pertainingtothisdataarereservedtoGEOTERRAIMAGEpty ltd.

14. appendix G


CONTENTS

1. iNtroDUCtioN 350

2. BaCKGroUND 350

3. oBJeCtive 350

4. proDUCt DeSCriptioN 351

4.1 land-Cover legend 351

5. overvieW oF tHe aUtoMateD MoDelliNG approaCH 351

5.1 Model portability to other Geographical areas and Sensors 352

5.2 Use of object Based Modelling 352

6. laNDSat iMaGery 352

6.1 landsat image Sourcing 354

6.2 landsat Data preparation 354

7. laND-Cover MoDelliNG 355

7.1 Spectral Modelling 355

7.2TerrainModifications 356

7.3 individual Frame Spectral Models 356

7.3.1 Spectral Modelling of Water 357

7.3.2 Spectral Modelling of Bare Ground 357

7.3.3 Spectral Modelling of tree / Bush Cover 357

7.3.4 Spectral Modelling of Grass 357

7.3.5 Spectral Modelling of Burnt areas 357

7.4SeasonallyDefinedSpectralLand-CoverperImageFrame 357

7.5 Generation of Full South african Spectral Class Mosaic 359

7.5.1 Grassland Corrections: inland 360

7.5.2 Grassland Corrections: Coastal 360


8. laND-Cover iNForMatioN GeNeratioN 360

8.1 Conversion from Spectral to information Classes 360

8.2 Creating Base vegetation land-Cover Classes 361

8.2.1 Fynbos 362

8.2.2 indigenous Forests 362

8.2.3 Wetlands 362

8.2.4 Nama and Succulent Karoo 363

8.3 Creating land-Use Classes 363

8.3.1 Cultivated lands 364

8.3.2 plantations 364

8.3.3 Mines 364

8.3.4 Built-up / Settlements 365

8.3.5 erosion 365

8.3.5.1 Degraded 365

9. aCCUraCy aSSeSSMeNt 366

9.1 Design and approach 366

9.2 results 368

9.3 analysis and Conclusions 369

9.3.1 Dense Bush / thicket 369

9.3.2 Woodland / open Bushland 370

9.3.3 Grassland 370

9.3.4 low Shrubland (Fynbos and other Sub-Classes combined) 370

9.3.5 Bare Ground 371

9.3.6 Conclusion 371

10. Data USe 371

14. appendix G


11. MetaData 371

appeNDix a 2013-14 South african National land-cover legend 374

APPENDIXBListofaccompanyingdocumentsandfiles. 380

APPENDIXC51xClassSeasonallyDefinedSpectralFoundationClassLegend. 381

appeNDix D visual representation of canopy cover density percentages per unit area. 382

APPENDIXE33xVerifiedLand-Cover/UseClasses. 383

PLEASE NOTE

Thisreportprovidesinformationonthespecificversionofthe2013-14GEOTERRAIMAGESouthAfricanNationalLand-Cover Dataset supplied to the South african Department of environmental affairs under the Dea / CarDNo project: SCpF002:”implementation of land Use Maps for South africa”.

the supplied dataset referred to in this report represents a subset of the full 2013-14 GeoterraiMaGe South african National land-Cover Dataset, in terms of supplied land-cover / land-use categories and associated class legend.

Use of this Dea / CarDNo data version is governed by the liCeNCe aGreeMeNt shown previously.


1. INTRODUCTION

land-cover is a key information requirement for a wide range of landscape planning, inventory and management activities, ranging from environmental resource management to telecommunication planning. the recent global availability of landsat 8 satellite imagery offered the opportunity to create a new, national land-cover dataset1 for South africa, circa 2013-14, replacing and updating the previous 1994 and 2000 South african National landcover datasets. the 2013-14 national land-cover dataset is based on 30x30m raster cells, and is ideally suited for ± 1:75,000 - 1:250,000 scale GiS-based mapping and modelling applications.

the dataset has been derived from multi-seasonal landsat 8 imagery, using operationally proven, semi-automated modelling procedures developed specifically for the generation of this dataset, based on repeatable and standardised modelling routines. the dataset has been created by GeoterraiMaGe (Gti) and is available as a commercial data product.

the production of a comparable 1990 land-cover product isalsoinprogressthatwillprovide,fortheveryfirsttime,a standardised and directly comparable set of national land-cover datasets that can be used to determine landscape change in South africa over a 20+ year period2.

2. bACKGROUND

the recent global availability of landsat 8 imagery (april 2013), was the primary catalyst behind the development of new, innovative automated land-cover modelling procedures, which allowed the creation of the 2013-14 South african National land-Cover dataset. primarily because landsat 8 offered a free, and regular supply of radiometrically and geometrically standardised, medium resolution, multi-spectral imagery, suitable for medium to large area mapping. Collectively, this offered an ideal opportunity for Gti to re-look at using automated land-covermappingtechniquesasamoreefficientalternativetoconventional,analyst-assistedper-pixelclassifiers,allowing rapid production of Standardised, yet informative land-cover information.

3. ObJECTIVE

the primary objective was to generate a new, national land-cover dataset for the whole of South africa, and to be able to release this as a commercial data product within ±1 year of image data acquisition to ensure that the dataset is up-to-date and current. in support of this primary objective was the requirement to develop operational procedures based on semi- automated (spectral) modelling techniques that would facilitate the rapid production of consistent, standardised land-cover information from multi-seasonal landsat 8 imagery.

1 Note that the term “land-cover” is used loosely to incorporate both land-cover and land-use information in the context of the Gti 2013-14 South african National land-Cover dataset.

2 the existing 1994 and 2000 South african National land-cover datasets were generated independently, using very different source image formats (i.e. paper versus digital), and end-product formats (i.e. 1:250,000 scale digital vector versus 1:50,000 digital raster), despite being derived from equivalent landsat imagery. See thompson MW, 1991. South african National land-Cover Database project. Data User’s Manual. Final report (phases 1, 2, & 3). CSir project report eNv/p/C 98136, February 1999; and thompson MW et al, 2001. Guideline procedures for National land-Cover Mapping and Change Monitoring. CSir Client project report eNv/p/C 2001-006, March 2001.

14. appendix G


4. PRODUCT DESCRIPTION

the 2013-14 South african National land-cover dataset produced by GeoterraiMaGe as a commercial data product has been generated from digital, multi-seasonal landsat 8 multispectral imagery, acquired between april 2013 and March 2014. in excess of 600

landsat images were used to generate the land-cover information, based on an average of 8 different seasonal image acquisition dates, within each of the 76 x image frames required to cover South africa. the land-cover dataset, which covers the whole of South africa, is presented in a map-corrected, raster format, based on 30x30m cells equivalent to the image resolution of the source landsat 8 multi-spectral imagery. the full dataset contains 72 x land-cover / use information classes, covering a wide range of natural and man-made landscape characteristics. TheDEA/CARDNOdatasetcontainsasummarised 35 x class legend. each data cell contains a single code representing the dominant land-cover class (by area) within that 30x30m unit, as determined from analysis of the multi-date imagery acquired over that image frame. the original land-cover dataset was processed in UtM (north) / WGS84 map projection format based on the landsat 8 standard map projection format as provided by the USGS3.ThefinalproductisavailableinUTM35(north)and (south), WGS84 map projections and Geographic Coordinates, WGS84.

4.1 Land-Cover Legend

the 2013-14 Sa National land-cover legend is aligned with SaNS 18774 South african National land-cover

ClassificationStandardintermsofclassdefinitionsandhierarchical format. the complete list of information classes that have been mapped and supplied as part of the Dea / CarDNo land-cover dataset, and the associated classdefinitionsaresuppliedinAppendix A.

Note that in recent CD:NGi5 national land-cover mapping projects, the Fao lCCS26land-coverclassificationsystemhas been adopted as an alternative to SaNS 1877, although this is not a national standard. However in September 2014 the Sa Geo7 Group’s land-Cover Community of practice (Cop) task team, which includes CD:NGi representation, proposed that the class detail and content of SaNS 1877 be retained / re-included within a proposed new Sa national land-cover standard, which will be based on the latest, internationally accepted land-Cover MetaData Language(LCML)classificationsystem(ISO1944-2).

5. OVERVIEW OF THE AUTO- MATED MODELLING APPROACH

automated modelling procedures offer signif icant advantages in terms of ensuring data standards, minimising processing time, allowing easy repeatability and facilitating accurate change detection; when compared to more conventional image mapping approaches where there is a greater reliance on individual image analyst knowledge and inputs. to this end, a series of automated image-based modelling steps were developed that utilise the seasonal dynamics associated with the broad landscape characteristics across South africa. these were then used

3 United States Geological Survey.

4SANS1877:SABureauofStandardsdesignatednationalland-coverclassificationstandardforSouthAfrica.

5 Chief Directorate: National Geospatial information

6 FoodandAgricultureOrganisation:LandCoverClassificationvs2

7 Sa Group earth observation


to rapidly produce a set of “foundation” cover classes that could be easily converted into more meaningful land-cover informationcategories,usingpre-definedgeographicmasks in the GeoterraiMaGe data libraries.

the “foundation” cover classes represent the basic building blocks associated with all landscape characteristics, namely water, bare ground, grass and tree-bush-shrub covertypes,witheachbeingdefinedintermsofseasonaloccurrence or permanence. these basic foundation cover classes represented the initial output from the automated modelling approach. the foundation cover classes are then converted into more conventional land- cover information classes, i.e. urban, forest plantation etc, as part of the post-automated modelling data processing steps.

the foundation cover classes are essentially “spectrally dependent” classes, since they are generated from automated modelling procedures that are based directly on the spectral characteristics associated with each image pixel over time (i.e. seasonal) and space (i.e. within an imageframe).Thefinalland-coverinformationclassesare referred to as “spectrally independent” classes since different cover classes can share similar foundation class spectral characteristics in a one-to-many type relationship. For example, the “bare ground” spectrally dependent cover classes could represent non-vegetated built-up urban areas, natural rock exposures, beach sand or a mine pit and tailings dump. Similarly the tree-bush classes could represent a natural vegetation cover, a timber plantation or a fruit orchard. the advantage of this approach is that the conversion of the initial, spectrally dependent foundation coverclassesintothefinal,spectrallyindependentland-cover information classes can be tailored to suit a variety of end-user information requirements; simply by using a different set of pre-determined masks and foundation class sub-divisions and amalgamations.

5.1 Model Portability to other Geographical Areas and Sensors

although model development was focussed on using

landsat 8 imagery within South africa, the same models have also been proven to work with equivalent success on landsat 8 imagery over sites in Sudan, Zimbabwe, Botswana, Namibia and Mozambique; as well as using comparable landsat 5 archive imagery over South africa; all of which indicate a high level of model portability. this should allow and support the production of directly comparable historical land-cover datasets for change detection, assuming of course that the required level of seasonal image coverage is available in the data archives.

5.2 Use of Object based Modelling

No attempt was made to include or use object-based modelling in the automated mapping process, primarily because the medium resolution format of multi-spectral landsat 8 imagery does not lend itself to this type of modelling; since pixel resolution typically precludes the identificationoftruelandscape“objects”incomparisonto high and ultra-high resolution image formats. the 30 metre landat resolution pixels typically represent a mix of land-cover characteristics rather than a pure cover surface, i.e. an urban pixel is typically a composite of building roofs, garden vegetation and / or road surfaces etc. However, object-based modelling may be useful approach for helping to generate 2nd level information classes from the primary level “foundation” class dataset, such as for example, separating water in rivers (i.e. natural)fromwaterindams(i.e.artificial),basedonsize,shape and context etc.

6. LANDSAT IMAGERY

the primary imagery source used in the generation of the 2013-14 Sa National land-cover dataset was 30m resolution landsat 8 imagery, acquired between april 2013 and March 2014. Within each image frame, a range of seasonal image acquisition dates were used (within the automated modelling procedures), to characterise the

14. appendix G


seasonal dynamics across the landscape in terms of the basic tree, bush, grass, water and bare foundation cover classes. Nine image acquisition dates per image frame, per complete seasonal cycle was taken as the optimal number of dates, based on landsat’s 16-day overpass schedule8. Unfortunately due to localised, prolonged cloud cover problems in some regions during this april 2013 - March 2014 period, this was not possible to achieve. in such cases, archival landsat 5 imagery was used as a substitute, but only if it was from a suitable seasonal period to compliment the landsat 8 data, and was ± 100% cloud free. Figure 1 illustrates the total number of image acquisition dates per image frame, whilst Figure 2 illustrates the location and number of landsat

5 images that were also used.

in summary, 76 x landsat image frames were required to provide complete coverage of South africa, Swaziland and lesotho. 616 x individual landsat images were used to model and produce the land-cover data, representing and average of 8 x acquisitions per image frame per year. of the 616 x images, 592 x images were from landsat 8 (i.e. 96 %) and 24 x images (i.e. 4 %) were from landsat 5. the landsat 5 images were however only used in 10 ofthe76imageframesdefiningthegeographicextentofthe land-cover dataset.

8 this seasonal window requirement, whilst improving the overall accuracy of interpretation and land-cover modelling, means that in practice thatbetween12-16monthsareactuallyrequiredtoacquiresufficientcloudfreeseasonalimageryforinputintothemodellingprocesses.this means that a new land-cover dataset could only be generated approximately every 2 - 3 years (minimum), since a shorter period may notprovidesufficientseparationbetweentheseasonalinputimagedates.

Figure 1: total number of image acquisition dates used in the land-cover modelling per image frame.


6.1 Landsat Image Sourcing

all landsat 8 and 5 imagery used in the 2013-14 land-cover data modelling was sourced from the on-line data archives of the United States Geological Survey (USGS, http://glovis.usgs.gov/). the data are provided in a precise geo-corrected UtM (north) / WGS84 map projection format, and was used “as-is” without any further geo-correction. a complete list of the image acquisition dates used in the modelling of the 2013-14 South african land-cover dataset (for both landsat 8 and 5) is contained in the accompanying spreadsheet (see Appendix b).

6.2 Landsat Data Preparation

all landsat 8 and 5 imagery was standardised to 16-bit,

Top-of-Atmosphere(ToA)reflectancevaluespriortoland-cover modelling using a generic modelling approach. the original USGS generated UtM (north), WGS84 map projection format was retained without change or modificationthroughoutallimage-framebasedmodellingprocedures. as far as possible only cloud free or image dates with limited cloud cover were used in the land-cover modelling (i.e. maximum ± 20% terrestrial cloud cover in any one date). any cloud affected regions were corrected by merging cloud affected toa corrected imagery with cloud free toa corrected data from preferably the preceding or following overpass date; so that as far as possible,thefinalcloud-freemergedimagerycompositeonly represented a maximum difference of ± 16 days. Cloud masks were created using either using conventional pixelclassificationprocedures(withinanalystdefinedsub-image areas), or spectral based modelling using generic

Figure 2: location of the 10 x image frames that required the use of archival landsat 5 imagery due to the limited availability of landsat 8 data during the period april 2013 to March 2014.

14. appendix G


thermalandbluelightreflectancethresholds9; depending on cloud characteristics. this was deemed acceptable in terms of minimising any changes in local vegetation cover growth changes. approximately 35% of the 616 x images used in the land- cover modelling were cloud-masked composites. No external atmospheric correction was applied to the image data.

7. LAND-COVER MODELLING

7.1 Spectral Modelling

Derived spectral indices, generated from the toa corrected imagery were used as the only inputs into the land-cover models. No original spectral (i.e. ToAreflectance)datawasusedasaninput,althoughcollectively, 5 out of the 8 available non-thermal spectral bands were used in the various spectral indices10. a standardisedsetofspectralindiceswereidentifiedfromwhich the required foundation cover classes, namely (1) tree dominated, (2) bush dominated, (3) grass dominated, (4) water and (5) bare ground could be modelled, using pre-determined, generic spectral threshold values. the generic threshold values associated with each indices and cover type were tested over several landscapes and seasonsbeforebeingconfirmedandacceptedassuch.

the spectral indices included both existing algorithms such as the Normalised Difference vegetation index (NDvi) and Normalised Difference Water index (NDWi), aswellasalgorithmsdevelopedin-housespecificallyforthe Gti land-cover modelling requirements. all models were developed in erDaS imagine © image processing software using the Model Maker function.

the modelling and generation of each foundation cover class was undertaken as a separate modelling exercise, i.e. water was modelled separately from bare ground, which was modelled separately from tree and bush coveretc.Thisapproachsimplifiedthemodellingstepsand facilitated easier desk-top quality control of outputs; compared to attempting to model all foundation cover typessimultaneouslywithinasinglemodelworkflow.Inmostcasesthefinalgeographicextentofaparticularfoundation cover class was generated from the combined use of several different spectral indices, since no single index was found to work well in all landscapes and all seasons.

all modelling was undertaken on an individual image frame-by-frame basis, where the original USGS supplied UtM (north) map projection format was retained as-is. this allowed the models to be adapted according to the number of acquisition dates and associated seasonal ranges available for each particular image frame. the modelmodificationsreflectedtheneedtostandardisetheoutputsaccordingtopre-defineddefinitionsofseasonalpermanence, for example, whether a foundation cover class occurred, i.e. in only one image date, in several, but not all image dates, or in all image dates. obviously the more image acquisition dates available per frame, and the wider the seasonal range, the more accurate the modelled interpretation of a particular cover class’s seasonal characteristics.

examples of the separate model outputs for bare ground and tree/bush foundation cover classes are shown in Figure 3. the examples show eastern pretoria, from within landsat 8 frame 170-078.

9 the spectral modelling approach could only be applied to landsat 8 imagery (if applicable) due to its enhanced spectral band range, compared to landsat 5.

10 (2) Green, (3) red, (4) red, (5) Nir, (6) SWir-1 and (7) SWir-2 landsat 8 spectral bands.


in some circumstances, modifications or additional modelling steps were used to account for regional landscape characteristics that could not be accurately modelled using the only generic models. For example, in some of the far western arid areas, additional tree-bushmodellingstepsweretakenusingaslightlymodifiedmodelling approach in order to better detect and separate between the sparse, low bush and shrub covers in these landscapes.

7.2 Terrain Modifications

Terrainbasedmodificationswereincludedwithinsomeof the foundation cover class modelling procedures (i.e. water, bare and tree-bush classes), in order to minimise seasonally induced terrain shadowing effects, using a combination of solar illumination and slope parameters.

these parameters were modelled independently using the 90m resolution SrtM dataset11, with outputs being re-sampled to a landsat comparable 30m resolution format, prior to being incorporated into the foundation cover class models. For example, solar illumination and DeM slope masks were used to minimise any spectral confusion between dark terrain shadow areas and comparable water bodyreflectancelevels.NotethatwhencreatingtheDEMderived parameters, such as shadow extents, modelling was based on the full range of seasonal solar azimuth and zenithanglesandnotaspecificdateorseason,inorderto better align with seasonal differences evident across the any given range of image acquisition dates.

7.3 Individual Frame Spectral Models

the basic modelling parameters used to create each of

Figure 3: Spectral indices derived model outputs for bare ground and tree bush dominated foundation cover classes, over eastern pretoria (frame 170-078).

11 although the 30m SrtM terrain dataset would have been more preferable it did not become publically accessible until after the completion of the bulk of the 2013-14 land-cover modelling had been completed.

14. appendix G


the “foundation” cover classes are described below. in each case the modelled output would typically represent a (qualitative) gradient of spatial densities and as well as temporal, i.e. seasonal occurrence. For example the bare ground model output consisted of a set of classes that represented a range of conditions from “pure bare” to “mixed bare / sparsely vegetated”, with each described in terms of seasonal permanent or temporary occurrence. the foundation water class modelling was the only exception to this, with the model output describing water extent in terms of only permanent or seasonal occurrence.

7.3.1 Spectral Modelling of Water

Water was modelled using a combination of 5 spectral indices, including a burn index, and DeM derived solar illumination and slope masks. the burn index was used to minimise any spectral confusion between dark burnt areasandcomparablewaterbodyreflectance.Thesolarillumination and slope masks were used to minimise any spectral confusion between dark terrain shadow areas andcomparablewaterbodyreflectance.Genericspectralwatermodelsweredevelopedforbothflat(standard)and hilly terrain, and for shallow water (pan) dominated landscapes.

7.3.2 Spectral Modelling of Bare Ground

Bare ground was modelled using a combination of 2 spectral indices, and DeM derived solar illumination and slope masks. the solar illumination and slope masks were used to minimise any spectral confusion between dark terrain shadow areas and comparable bare ground reflectance.Theoutputrepresentedagradientofbareground from pure bare to sparsely vegetated cover.

7.3.3SpectralModellingofTree/BushCover

tree and bush cover was generated within the same spectral model. Both woody covers were modelled using a combination of 4 spectral indices, and DeM derived

slope mask. the slope mask was used to minimise any spectral confusion between dark terrain shadow areas and comparable woody cover in two of the 4 spectral indices. “trees” refers to dense, typically tall woody cover, such as natural and planted forests, dense woodland and thickets, etc; whereas “bush” refers to more open, often lower mixed tree / bush communities such as typical bushveld or open woodland. the spectral modelling procedure also separated woody cover spectral classes that showed similar characteristics to other vegetated covers such as crops,sportsfields,golffields(especiallyifirrigated),fromthose that didn’t. an additional “desert” tree / bush model was developed as an extra model to be run in addition to the standard model in more arid areas, in increase the representation of bush cover in these regions where the modelling thresholds used in the standard model were not sensitive enough.

7.3.4 Spectral Modelling of Grass

Grass cover was modelled using only a single spectral indices. No DeM derived solar illumination and slope maskmodifierswereused.

7.3.5 Spectral Modelling of Burnt Areas

Burnt cover was modelled using only a single spectral indices. No DeM derived solar illumination and slope maskmodifierswereused.

7.4 Seasonally Defined Spectral Land- Cover per Image Frame

the modelled outputs for each foundation cover class were combined into a single composite dataset for each frame, in a pre-determined hierarchical order, in order thatthefinaloutputcoverclasscombinationsreflectedthe dominant and sub-dominant cover types in terms of seasonal occurrence. For example:

• water in all dates (permanent),

• water in many dates (seasonal),


• tree-dominated in all dates,

• tree-dominated in many but not all dates plus bare ground in one date, etc

Figure 4 shows the output after all the separate foundation land-cover classes have been combined into a compositesetofseasonallydefinedfoundationland-coverclasses. the illustrated area is the same eastern pretoria area as shown in Figure 3 previously. this interim land-coverproductconsistsof51xseasonallydefinedfoundation cover classes, as listed in Appendix C. these seasonally combined foundation cover classes are still essentially “spectrally dependent” classes, since they have been generated using automated modelling procedures, on only the spectral characteristics associated with each image pixel over time (i.e. seasonal) and space (i.e. within an image frame).

Note that in addition to the basic tree, bush, grass, water and bare ground foundation classes already described, firescarextentandoccurrencewasalsomodelledandincorporatedintothefinalcombinedfoundationcovercomposite classes, in order to provide an indication of whetherornotfirewasafactorinthemodelledseasonalcoverprofiles,i.e.asingledate,baregroundareacouldbetheresultofawildfireeventthatthenre-vegetatedlater in the season.

the spectrally dependent and seasonally def ined foundation cover classes in this base dataset represent the “building blocks”, from which more information focused land-cover classes can be derived, especially if combined with ancillary datasets such as vegetation or land-use maps to facilitate further class sub-divisions or area-basedclassmodifications.Theentirespectralmodellingprocedure is summarised graphically in Figure 5.

Figure4: Interimspectralland-coverproductaftercombiningalltheseasonaldefinedfoundationcoverclasses,(easternPretoria,frame170-078)

14. appendix G


7.5 Generation of Full South African Spectral Class Mosaic

once the spectral modelling for the individual frames had been completed, the next step was to merge these into a single, seamless coverage for the whole of South africa. achieving this involved several processing steps. Individualimageframedatasetswerefirstedge-cleaned

to ensure that no frame-edge pixel anomalies would be carriedoverintothefinalcountry-widemosaic.Cleanedimage frames were then re-projected, were necessary, to a standardised UtM35 (north), WGS84 map projection; since the majority of image datasets covering South africa were sourced from the USGS data archives were in this UtM zone.

individual image frames were merged into a single, country-wide spectral foundation class composite, based on an analyst determined sequence of image border overlap rules in order to ensure the best possible seamless

integration of the individual frames. this approach was necessary since some image-to-image edge differences do occur between image frames where spectral foundation classeshavebeengeneratedfromsignificantlydifferentinput acquisition dates, simply as a result of image data availability.

although the standardised modelling approach used to generate the es supported generation of spatial and content comparable land-cover data in adjacent image frames (if both framesutilised comparable input imageacquisitiondates), if therewasasignificantdifferencein the number or seasonal range of input images, then adjacent images often showed edge matching differences. in such cases, the adjacent images were often spatially comparable, but not necessarily content comparable. in these cases, the edge matching process was set-up, as far as possible so that the most prominent and spatially extensive landscape pattern was maintained across the image borders, in order to provide a more visually seamless image frame overlap.

Figure 5: Summary of the full spectral land-cover modelling process


Note that for water (“seasonal” and “permanent”) and bare ground (“pure bare, all dates”) country-wide coverages, the maximum extent of each class, from all images,wastransferredacrosstothefinalcountry-widespectral class composite; regardless of any frame overlap edge matching applied to other spectral classes.

7.5.1 Grassland Corrections: Inland

persistent cloud cover problems on several image acquisition dates during the 2013-14 wet season period over the Highveld region resulted in the unavailability of usable, early wet season imagery for modelling inputs. this resulted in the standard spectral models under- estimating the grassland extent. in order to correct this, an alternative grassland model was developed, based on a single, rather than multiple, image acquisition date. the optimal single date representing the best, dry season image for visually separating green woody vegetation from senescent grassland vegetation, with minimal burn scars. Forimageframescoveringareasofsignificanttopographicalrelief, such as over the escarpment, the single- date grass corrective model included use of solar illumination and slope parameters to improve modelling outputs. For standardisation purposes, the inland grassland corrective model was applied to all image frames overlapping or contained within the SaNBi grassland Biome boundary. the resultant grassland extents for each image frame where then merged into a composite geographical mask and integrated back into the original county-wide spectral class dataset to improve the overall grassland delineation.

7.5.2 Grassland Corrections: Coastal

in some high rainfall coastal regions, the standard spectral models were unable to accurately distinguish between grassland and woody cover, where the grass biomass was exceptionally high as a result of high local rainfall during the 2013-14 wet season, even if suitable image acquisition dates were available. this was especially the case around pondoland (e Cape) and Zululand and Maputoland (KZN). to correct this another grassland correction model was

developed with alternate spectral thresholds to better delineate the correct grassland extent in these areas. the resultantframespecificoutputswerethenmergedintoa composite geographical mask and integrated back into the original county-wide spectral class dataset to improve the overall grassland delineation.

8. LAND-COVER INFORMATION GENERATION

Multi-seasonal landsat imagery has been used to generate a comprehensive set of spectrally-based outputs that describe the seasonal characteristics of tree, bush, grass, bare and water cover characteristics within each image pixel, as represented by the 51 x spectral foundation classes. these were converted into more meaningful and informative land-cover and land-use information classes, aligned with the SaNS 1877 South african National land- cover Classification Standard (see Section 4.1). For example a pixel may be described as having a detectable tree cover on more than one image acquisition date, but not all dates; and grass covered for several but not all dates. this would then be interpreted as an open, deciduous tree covered woody community, (i.e. woodland), where the grass cover became evident when the tree canopy showed less than maximum foliage cover.

8.1 Conversion from Spectral to Information Classes

the creation of both land-cover and land-use information classes involves the recoding and re-grouping of selected spectral foundation classes either in a controlled or uncontrolled geographic area. in a controlled re-coding, independently sourced and / or generated geographical masksareused todefine the locationandextentofthe recode process. this is required because in many situations different landscape features can be represented by similar spectral characteristics. For example a

14. appendix G


spectrally modelled area of permanent bare ground (i.e. spectral characteristics are representative of a non-vegetated surface in all image acquisition dates), could be representative of a beach, a mine, or a large building. By using an appropriate geographic mask (essentially an independentlydefinedorsourcedpolygoncoverage),itis then possible to code all bare ground pixels within the mask to, for example, mining, whilst allowing all other bare areas outside the mask to represent other bare ground surfaces, which can then be further re-coded as and where required.

the key to this process is that the spectral modelling definestheexactfootprintofthespectralcoverclass(inauser independent manner), whereas the geographic mask definestheinformationcontent(basedonuserexpertiseand knowledge). this approach overcomes problems where spectral characteristics are shared between two or more separate land-cover/use classes, which is

self-evident in many image-based mapping applications. Figure 6 illustrates the various procedural pathways that can be used to convert the spectral foundation classes into land-cover / land-use information classes.

8.2 Creating base Vegetation Land-Cover Classes

the natural vegetation base classes, namely thicket / dense bush, woodland / open bush, grassland, low shrubland and bare, non-vegetated were created by merging selected groups of spectral classes. Generally, selected spectral classes were merged into the required vegetation cover classes without geographical control, using the original spectral class extent and distribution ‘as-is’. However in some regions, and for some spectral classes it was necessary for the re-coding and merging process to be geographically controlled, and only allow the recode

Figure 6: processing pathways for converting spectral foundation classes to land-cover / land-use information classes, based on seasonal permanence and shared spectral characteristics.


procedure within pre-determined geographical regions. Controlled geographical recoding (i.e. masking), used pre-selected SaNBi bioregion12boundariestodefinewhereandhowselectedspectralclassesweremodified.Notethatthefinalvegetativeland-coverclassboundarywasalwaysdefinedbytheoriginalspectralfoundationclass boundaries contained within the SaNBi bioregion boundary, and not by the bioregion boundary itself, other than for the Fynbos low shrubland sub-class.

8.2.1 Fynbos

the sensitivity of the automated spectral models was insufficienttodeterminethetransitionbetweenstructurally similar Fynbos low shrub communities and non-Fynbos low shrub communities, except for the boundary with the structurally different Namaqualand (SKn and SKs) and Knersvlakte (Skk) Bioregions. in these areas the boundary between the Fynbos and non-Fynbos lowshrubcommunitieshasbeendefinedonthebasisofthe original spectral class boundaries13.

8.2.2 Indigenous Forests

the indigenous forest land-cover class was generated in a similar manner to the previously described base vegetation classes, although the geographical masks used to control the recoding of the selected spectral foundation classes were derived from a combination of independent datasets. these included the SaNBi forest

12 Sourced from “the vegetation of South africa, lesotho and Swaziland”. eds Mucina l and rutherford MC. 2006. Strelitzia 19. South african National Biodiversity institute: associated digital GiS database. Note that SaNBi Biome and Bioregion boundaries, whilst generically applicable for 1:250,000 digital map scales, is often considerablymoreaccurateonalocalscale,evendowntothenearest100minsomecases (refer Section 4.1.2, Mapping Scale. p 15)

13 the is primarily because satellite imagery such as landsat, with its combination of medium spatial resolution and limited spectral capabilities (comparedtohyper-spectralsensors),isprimarilyresponsivetostructuralvegetationcharacteristicsratherthancommunityfloristics.accurate delineation of Biome-related boundaries, especially if the boundary is a transition zone, rather than distinctborder, is at best, challengingwithimagerysuchasLandsat.Especiallyiffloristicdifferencesplayagreaterrolethanstructuraldifferencesindefiningtheborder. this is likely to be the case between the Fynbos, Succulent Karoo, Nama Karoo and surrounding Grassland and Desert Biomes.

14 the 30m resolution SrtM terrain data for africa was available for use at this time within the 2013-14 Sa land-Cover project, despite not having been available for the initial spectral modelling phase.

15 Normalised Difference vegetation index (NDvi)

biome boundaries, higher detail regional masks created in-house in support of previous provincial-level land-cover mapping, shadow and altitude terrain parameters modelled from the higher resolution 30m resolution SrtM DeM terrain data14, and NDvi15 spectral datasets. the NDvi dataset represented a nation-wide summary of the seasonal maximum NDvi recorded in each pixel, calculated from all image acquisition dates used in the spectral land-cover modelling. Note that separate models were required to recode the different Forest vegetation types (i.e. Northern afromontane, Sand, Scarp etc), with each one being based on different geographical mask combinations and associated data thresholds, in order to correctly model the forest patch distributions in these regions.Thefinal,modelledforestextentswerethenintegrated into the base vegetation land-cover classes.

8.2.3 Wetlands

the wetland class was created using independent and separate modelling routines that did not include reference to, or recoding of any of the original 51 x class spectral foundation classes. Depending on the available image acquisition dates per frame, wetland extent was modelled using either a dual or single image acquisition date process. the single date approach used the best “wettest” image date, whereas the dual date approach was based on the difference between wettest and driest image dates whereas the single (wettest) date modelling was only used as an alternative approach if the available acquisition dates

14. appendix G


did not provide suitable seasonal differences for the dual date, wet-dry date approach. Both modelling approaches only used inputs from multiple spectral indices, and not use the 51 x spectral foundation cover classes as used in the other vegetation class generation. a combination eight indices were used in the dual-date approach, with each being applied to both dates; whereas only four spectral indices were used in the single, wettest date approach. Both modelling approaches generated a dataset that represented the likelihood of wetlands occurring in the landscape, within which a threshold was determined on a frame-by-frame basis in order to best represent actual wetland extent.

the single (wettest) date modelling approach used terrain shadow masking to minimise spectral confusion with non-wetland regions, whereas the dual-date approach used morecomplexslopeandfloodplainmodellingtolimitthetopographic areas within which wetlands were likely to occur. Both approaches used the 90m SrtM terrain data as the source for topographic modelling.

the slope and floodplain model generated potential floodingzones,basedonpotentialfloodheights,whichwere taken to be indicative of areas most likely to be wet, and thus contain the highest probability of having wetlands.Thefloodheightswerecreatedbyestimatingthewater expansion level in river channels against a modelled stream surface height layer, (in comparison to actual terrain surface). the modelling provides an indication of potential lateral water extent at a range of heights above the stream base height in relation to the stream channel profileatthatlocation,irrespectiveofflooddurationortime periods. the modelling does not take account of the surfaceroughnessorflowblockagewhichmayresultinadditionallateralextensionoffloodwatersthroughbackupetc. the accuracy of the output is directly dependent on the scale, resolution and detail contained within the DeM. Themodelledpotentialfloodingzoneswerethenusedtorestrict the dual-date wetland modelling where wetlands are most likely to occur.

Note that in some localised areas, the results of both single

and dual-date wetland modelling approaches required post-modelling edits to improve the wetland delineation, especially if wetlands were not showing saturation or significant (vegetation)flushes. Insuchcases,analystassistedconventionalunsupervisedclassificationswereused to locally improve the modelled wetland output.

the modelled wetland extents were then integrated into the base vegetation land-cover classes.

8.2.4 Nama and Succulent Karoo

the Nama and Succulent Karoo categories and associated sub-classesareincludedasaresultofaspecificDEAinformation requirement in the Dea / CarDNo land-cover data product. Since the sensitivity of the automated spectral models was insufficient to determine these Biomeboundaries,theyhavebeendefinedsolelyontheSaNBi Biome boundaries, as per the previous Fynbos explanation.

However within these SaNBi boundaries the original level of modelled land-cover detail has been retained, in terms of forest, dense bush, open bush, low shrubland, grassland and bare ground sub-classes. this is because each Biome can contain a wide range of structural vegetation types, over and above the dominant structural form. For example, localised patches of thicket, tall bush and grass-like structural communities all occur within the general low shrub matrix that comprises the Nama Karoo Biome, according to local species composition, terrain location andfirehistory,etc

8.3 Creating Land-Use Classes

land-use classes such as cultivated lands, forest plantations, mines and settlements were all derived as separate modelling procedures, typically using independently sourced and generated geographical masks to control where and how the original 51 x seasonally defined,spectralfoundationclasseswerere-codedandmodifiedintothefinalland-useclasses.


8.3.1 Cultivated Lands

the Department of agriculture, Forestry and Fisheries’s (DAFF)publicdomain2013nationalfieldboundarydataset(captured from 2.5m resolution, pan-merge Spot5 imagery), was the initial source of the cultivated land-use classes. this existing dataset was updated nationally to represent the latest cultivated land patterns present on the latest 2014 landsat 8 imagery used in each image frame. this typically involved capturing the distribution of new,commerciallycultivatedfields,especiallycentrepivotirrigationunits,althoughallfieldtypes,wereincludedin the 2014 updating process, including subsistence cultivation areas, if clearly apparent.

in addition, the location, extent and distribution of all commercial pineapple and sugar cane crops were mapped from the same 2014 landsat 8 imagery, to increase the levelofcropspecificsub-classdetail,in-linewithseveralof the previous provincial land-cover datasets generated by GeoterraiMaGe.

all new cultivated lands (and crop types) mapped from the 2014 landsat 8 imagery were captured manually using on-screen photo-interpretation techniques; since this facilitated and maintained similar interpretation accuraciestothesourceSPOT5derivedfieldboundarydata.Furthermoreitalsoeliminatedpost-classificationeditingnecessitiestypicallyassociatedwithfieldboundary(asopposedtocroptype),classificationattempts.

Thefinalcultivatedlandboundarieswereused‘as-is’todefinethefinalgeographicalextentofallthecultivatedland-use classes incorporated into the land-cover dataset. No use of, or reference to, the original 51 x class, seasonallydefined,spectralfoundationclasseswasmade.

8.3.2 Plantations

Forestry plantations were derived from selected classes from within the 51 x spectral foundation cover dataset, which were re-coded and re-grouped within controlled geographical areas, using independently sourced and generated geographic masks. the plantation area masks were sourced from several in-house provincial mapping projects, updated nationally to the plantation extent visible on the 2014 landsat imagery.

Notethat thefinalplantationclassboundarieswerealwaysdefinedbytheoriginalspectralfoundationclassboundaries contained within the geographical masks, and not by the geographical mask boundaries themselves.

Thegeographicalmasksusedtodefinetheareastoberecorded for the clear-felled plantation sub-class were only mapped off the latest 2014 landsat imagery, in order to ensure that the representation and interpretation was current.

8.3.3 Mines

Mines were derived from selected classes from within the 51 x spectral foundation cover dataset, which were re-coded and re-grouped within controlled geographical areas, using independently sourced and generated geographic masks. the mine area masks were sourced from previous in-house, provincial mapping projects and national 1:50,000 scale map datasets16, all of which were then updated nationally to the mine activity extent visible on the 2014 landsat imagery. Note that mine water sub-classes were generated by identifying water classes that werelocatedwithinthefinalminegeographicalmasks.

16 Geographic masks from the national 1:50,000 scale topographic digital map data were created by extracting the relevant vector features and buffering them by 1 x landsat pixel extent. these were integrated with the in- house provincial mine masks, before updating to 2014 geographic extents off the 2014 landsat 8 imagery.

14. appendix G


Note that major road and rail features were excluded from the mine area footprint if it intersected the mine mask.

Notethatthefinalmineclassboundarieswerealwaysdefinedbytheoriginalspectralfoundationclassboundariescontained within the geographical masks, and not by the geographical mask boundaries themselves.

8.3.4Built-Up/Settlements

Built-up areas generated independently of the spectral foundation cover dataset. the primary source for the cover class were several, internally developed GeoterraiMaGe urban map products and associated databases. these provided detailed information on the extent, distribution and land-use for all settlements nationally.ThiswasverifiedinruralareaswiththenationalDwelling Frame dataset, available from StatS Sa17. this information was then spatially re-modelled to represent built-up area outlines, which were then further updated and corrected nationally, using the 2014 landsat imagery for visual reference. this was especially so in terms of rural village outlines.

the independently generated Built-up area class boundarieswereused‘as-is’todefinethefinalgeographicalextent of the settlement patterns incorporated into the land-cover dataset. No use of, or reference to, the original 51xclass,seasonallydefined,spectralfoundationclasseswas made.

8.3.5 Erosion

the erosion (donga) class was derived from selected classes (representative of bare ground) from within the 51 x spectral foundation cover dataset, which were re-coded and re- grouped within controlled geographical

areas, using independently sourced and generated geographic masks. the erosion area masks were sourced from previous in-house, provincial mapping projects and other national erosion datasets, all of which were then updated nationally to represent the current extent of major donga’s visible on the 2014 landsat imagery. Note that as a result of spectral modelling sensitivities, and the need to be able to

separate the bare ground within donga features from the surroundingnon-erodingareas,thefinalmodelledextentoferosionfeaturesissignificantlybetterrepresentedbothspatially and numerically in the wetter, more vegetatively lush regions of the country, where the non- vegetated erosion surface are significantly different from the surrounding vegetation cover (i.e. bushveld and grassland regions). Donga feature detection in the drier more arid region is not as accurate.

Aswithpreviousland-useclasses,thefinaldonga/erosionclassboundarieswerealwaysdefinedbytheoriginalspectral foundation class boundaries contained within the geographical masks, and not by the geographical mask boundaries themselves.

8.3.5.1Degraded

as part of the Department of environment affairs commissioned version of the 2014 GeoterraiMaGe South african land-Cover dataset, and additional sub-class,defining‘Degraded’areaswasalsogenerated.Thissub-class is however not part of the standard land-cover dataproduct.Degradedareasaredefinedintermsofthisdatasetasareasofsignificantlyreducedvegetationcover compared to immediately adjacent pristine or semi- pristine natural areas. Degraded areas were modelled from selected classes from within the 51 x spectral foundation cover dataset, which were re-coded and

17 Statistics South africa (http://www.statssa.gov.za/)


re-grouped within controlled geographical areas, using independently generated geographic masks. the basic degraded area mask was created from image derived, seasonally summarised NDvi maximum and standard deviation datasets. the extent of this mask allocation was limited to areas outside of formally protected (conservation) areas18, within non-arid Biome regions19, on terrain slopes < 21 degrees20, and not overlapping majorroads,dryriverbeds,beachesanddunefields21.

this modelling approach, whilst relevant in terms of practicalavailableinputdata,excludedtheidentificationof cover rich, species poor degraded areas, as well as degraded low vegetation areas within naturally occurring low vegetation regions. in the latter case alternate modelling options were considered, based on buffered threshold distances around settlements and mines etc, but the generic outputs were not considered universally reliable nor accurate.

Aswithprevious land-useclasses,thefinaldegradedarea boundaries were always defined by the original spectral foundation class boundaries contained within the geographical masks, and not by the geographical mask boundaries themselves.

9. ACCURACY ASSESSMENT

the 2013-14 South african National land-Cover dataset hasbeenverifiedintermsofmappingaccuracyinordertoprovideameasureofenduserconfidence indatause. the satellite image generated land-cover / land-use informationwasverifiedvisually,aspartofadesk-top

only procedure, against equivalent date, high resolution imagery and photography in Google earth ©. accuracies are reported using industry standard error (confusion) matrices, and include producer, User and Kappa values.

9.1 Design and Approach

Sample points for verification were selected using two separate approaches that considered the validity of statistical representation, the spatial resolution of landsat imagery in relation to landscape features, the national reporting frame extent, and the structure of the mapped land-cover information. Note that 33 x land-cover/land-useclasseswereverified,representinginsome cases primary rather than sub-class land- cover or land-use characteristics. Were necessary these were then amalgamated into the broader Dea / CarDNo land-cover / use classes for statistical reporting. For example,water,minesandplantationswereallverifiedat the primary level since sub-class detail was linked to shorttermtemporaleffectsthatcouldnotbeverifiedonthe single Google earth imagery.

Selection of samples representing potential commission errors was achieved by selecting ± 150 x samples for eachofthe33xclassestobeverified,givenatotalof±3500 points. points were selected randomly across the full national land-cover dataset. typically ± 100 samples were selected from thematic class units < 100 ha, and ± 50 samples were selected from thematic class units > 100 ha, in order to ensure a balanced representation of thematic unit areas. in all cases, the actual sample point represented the centre of the selected thematic unit, with placement set to exclude areas within 50 m of thematic unit boundary, to minimise class edge effects.

18 AsdefinedintheDEAProtectedAreas2014PACAdatabase.

19 Toensurethatdegradedareaswereclearlyidentifiableashavingalowervegetationcoverthanthesurroundingnon-degradedareas,unlessthey were in located within former tBvC state boundaries and closely associated areas of dense rural settlements.

20 to exclude natural rocky slopes and cliff faces with similar non-vegetated spectral characteristics.

21 to exclude other land-cover and land-use features with similar non-vegetated spectral characteristics.

14. appendix G


Selection of samples representing potential omission errors was achieved by selecting approximately ± 2500 samples, without reference to class type, on initially a stratifiedapproachusingtheimageframeextents,andthen randomly, within the selected frames, randomly across the original spectral foundation class dataset (as opposedtothefinalland-coverdataset).Theboundariesof the spectral foundation classes were just used as a sampling frame. points represented the centre of the selected thematic spectral unit, with placement set to exclude areas within 50 m of thematic unit boundary, to minimise class edge effects. this approach was deemed appropriate since it ensured that the sample points at leastreflectedthespatialvariabilityofthemulti-dateinput imagery, which in turn reflected the seasonal landscape characteristics. approximately 1000 samples were selected from thematic class units between 100 - 200 ha in size, ± 1000 samples from thematic class units between 200 - 500 ha, ± 400 samples from thematic class units between 500 - 1000 ha, and ± 100 samples from thematic class units > 1000 ha, in order to ensure a balanced representation of thematic unit areas.

NotethattheresultspresentedbelowrepresenttheDEA/

CARDNOland-coverlegendcontentandnotthefullland-coverlegendfromwhichithasbeenderived.

Note that no map accuracy statistics are provided at the Biome level, other than for the Fynbos: low Shrubland class,sincetheBiomedefinedboundarieshavebeenderived from fixed, independent data and simply overlaid on the spectrally modelled structural vegetation categories. the low Shrubland: Fynbos class is retained within the accuracy assessment since its delineation has, inpart,beendefinedonmodelledspectralcharacteristics.

Note that no map accuracy is provided for the Degraded class due to the diff iculty of reliably determining, visually,thatthevegetationataspecificsamplepointissignificantlylowerthanthatinsurrounding,undisturbednaturalvegetationareas,withoutin-fieldobservationoverwider areas. it is suggested that the calculated accuracy for bare ground and erosion are taken as indicative of the degraded mapping accuracy, since the degraded class is a modelled derivative of these classes.

Figure 7 illustrates the number and distribution of sample points across the full country extent.

Figure 7: Number and distribution of 6415 x sample points across the full country extent.


9.2 Results

the map accuracy results for the 2013-14 South african National land-Cover dataset, modelled from multi-seasonal landsat 8 imagery, based on the modif iedlegendformatrepresentingtheDEA-CARDNOinformationrequirements, are as follows:

the overall map accuracy for the 2013-14 South african National land-Cover DEA-CARDNO format dataset, modelled from multi-seasonal landsat 8 imagery is 82.53%, with a mean land-cover / land-use class accuracy of 88.36 %. this has been determined from 6415 sample points representing 17 x Dea-CarDNo land-cover / use classes. the Kappa index value of 80.87 indicates that these results are very unlikely to be the result of chance occurrence.

a breakdown of individual class accuracies is provided below

Note:classnumbersrefertheclassnumbersinthedigitalland-coverdata,whichincludeadditionalsub-classinformationrelatingtotheFynbos,NamaandSucculentKaroobiomegroupings.

14. appendix G


Theseresultsexceedtherequirementsasdefinedinthe terms of reference, that “a minimum map accuracy of70%istobeachievedatthe90%confidencelimits,witha minimum 70% Kappa index value ... with a minimum of 30 x samples per class ... (and) a target of 1000 x samples in total”.

Asignificantmajorityofindividualland-cover/land-useclasses achieved User and producer accuracies of 70% or more (see yellow highlights). Fourteen out of seventeen (14/17, 82%) classes achieved a Users accuracy greater or equal to 70%. thirteen out of seventeen (13/17, 76%) classes achieved a producer accuracy greater or equal to 70%.

producers accuracy (omission error) represents how well referencepixelsofthetrueland-coverareclassified.Users accuracy (commission error), represents the probabilitythatapixelclassifiedintoagivencategoryactually represents that category on the ground.

producers accuracy therefore represents the accuracy of the sample points whereas the Users accuracy represents what the data user will experience when using the land-cover data. the Kappa index is measure of statistical chance, with higher values (>0.7) typically representing repeatable and reliable results.

the contingency matrix showing the per-cover class breakdown of these map accuracy statistics is shown in Appendix E.

9.3 Analysis and Conclusions

overall the map accuracies for the 2013-14 South african National land-Cover dataset are very good and indicate that the dataset is accurate and that data users can have confidenceintheinformationcontainedwithin.Thereare

however some individual classes, with less than 80% User and / or producer accuracies that need further analysis and comment. all of these, bar one, are associated with spectrally modelled base vegetation classes, rather than the post-modelling derived land-use classes.

9.3.1DenseBush/Thicket

the Dense Bush / thicket class achieved a 83.64% producers accuracy but only a 53.86% Users accuracy. this appears to be primarily result of confusion with Woodland / open Bush (76/428 samples), Grassland (51/428 samples), and low Shrubland22 (45/428 samples). there is no clear regional pattern associated with this misclassification.

Within the Woodland / open Bush confused samples, 52ofthe76sampleshadawoodycover≥55%,and26of the 76 samples had awoody cover≥±65%,which may explain this inter-class confusion, since the (Google earth) observed canopy cover amounts could be considered close to the transition to dense bush cover. TheconfusionwithobservedGrasslandareasisdifficultto explain since there is very little woody cover observed in these samples. a possible explanation may be that localised areas of extremely high grassland biomass are (spectrally) being confused in the modelling process with woody cover. the Dense Bush / thicket confusion with low Shrubland may associated, in some cases, with the tall woody cover density of the low Shrubland23, since 24/45 of the incorrect sample sites have an observed woodycoverof≥±15%,15/45haveanobservedwoodycoverof≥±25%,9/45haveanobservedwoodycoverof≥±35%,and5/45haveanobservedwoodycoverof≥±45%.

if all the Dense Bush / thicket samples are grouped into tall woody versus non-woody (or low woody shrub)

22 low Shrubland Fynbos and other sub-classes combined.

23 the observed woody cover percentage for low Shrubland (Fynbos and other) as indicated in the accuracy spreadsheet represents the tall woody cover and not the general low matrix of woody shrubs.


cover24, then 349/428 (81%) samples have a tall woody canopycover≥±15%,regardlessofobservedland-covertype, 331/428 (77%) samples have a tall woody canopy cover≥±25%,323/428(75%)sampleshaveatallwoodycanopycover≥±35%,310/428(72%)sampleshaveatallwoodycanopycover≥±45%,and197/428(46%)sampleshaveatallwoodycanopycover≥±65%.

9.3.2Woodland/OpenBushland

the Woodland / open Bushland class achieved a 60.84% Users accuracy and producers accuracy of 54.13% Users accuracy. this appears to be primarily result of confusion with Grassland (59/452 samples), and low Shrubland25 (80/452 samples). there is no clear regional pattern associatedwiththismisclassification.

TheconfusionwithGrasslandisdifficulttoexplain,otherthan as a result of modelling and interpretation errors associated with lower woody canopy cover densities and associated spectral characteristics, with 14/59 (23%) of the observed Grassland samples having a woody cover of ≥15%.SimilarconfusionwasfoundwithLowShrubland,with 17/80 (21%) of the observed low Shrubland samples havingantallwoodycoverof≥15%cover,irrespectiveof the background low woody shrub matrix.

if all the Woodland / open Bushland samples are grouped into tall woody versus non-woody (or low woody shrub) cover, then 320/452 (70%) samples have a tall woody canopycover≥±15%,regardlessofobservedland-covertype, 289/452 (63%) samples have a tall woody canopy cover≥±25%,284/452(62%)sampleshaveatallwoodycanopycover≥±35%,221/452(48%)sampleshaveatallwoodycanopycover≥±45%,and80/452(17%)sampleshaveatallwoodycanopycover≥±65%.

9.3.3 Grassland

the Grassland class achieved a 84.56% User accuracy but only a 69.82% producers accuracy. this appears to be primarily result of confusion with Woodland / open Bush (82/1004 samples), and low Shrubland26 (55/1004 samples). there is no clear regional pattern associated withthismisclassification.

TheconfusionwithWoodland/OpenBushisdifficultto explain, other than as a result of modelling and interpretation errors associated, since the majority of observed Woodland / open Bush sample sites all have tall woodycoverdensities≥±55%.Howeveronthemajorityof observed low Shrubland sites, the woody cover density was less than ± 5 % cover, irrespective of the background low woody shrub matrix, which is probably representative of a transitional zone.

if all the Grassland samples are grouped into tall woody versus non-woody cover, then 813/1004 (80%) samples haveatallwoodycanopycover≤±15%,regardlessofobserved land-cover type and 711/1004 (70%) samples haveatallwoodycanopycover≤±5%,regardlessofobserved land-cover type.

9.3.4 Low Shrubland (Fynbos and Other Sub- Classes combined)

the low Shrubland classes for Fynbos and other respectively achieved User accuracies of 79.64% and 70.59%; and producers accuracies of 93.31% and 61.82%. these two classes have been assessed in combination since they are only differentiated on the basis of the overlaid SaNBi Biome boundary. in both cases confusion was primarily with Grassland (142/858 samples).

24 Note that Forest plantations do not have a woody canopy cover value recorded for either observed or mapped plantation sample sites, so it is likely that the accuracy of the modelled extent of dense woody cover is actually higher, since there are 8 x Dense Bush / thicket sample points that were observed to be plantation Forests.



14. appendix G


there is no clear regional pattern associated with this misclassification.

if all the low Shrubland samples are grouped into (tall) woody versus non-woody cover, then 791/858 (92%) samples have a tall woody canopy cover less than ± 15%, regardless of observed land-cover type and 691/858 (80%) samples have a tall woody canopy cover less than ± 5%, regardless of observed land-cover type.

9.3.5 Bare Ground

the Bare Ground class achieved a User accuracy of 73.54% and a producers accuracy of 77.54%. this appears to be primarily result of confusion with Grassland (26/378 samples), and low Shrubland27 (65/378 samples). there is no clear regional pattern associated with this misclassification.TheconfusionwithLowShrublandisdifficulttoexplain,otherthanasaresultofmodellingand interpretation errors associated, since the 57 / 65 observed low Shrubland sample sites all have tall woody cover densities between ± 5 - 55%, with the majority being between ± 15 - 25%. However these percentages, as indicated, only refer to the tall woody cover component, and not the low woody shrub component, which may have been very sparse. the confusion with Grassland may be due a comparable, overall low vegetation base cover, although unlike with the observed low Shrubland samples, the observed Grassland samples all had, bar one, zero % tall woody cover densities.

9.3.6 Conclusion

in conclusion it would seem that the majority of mapping errors associated with the base natural vegetation cover

classes can be broadly associated with transitional vegetation conditions, represented by structural vegetation gradients, within which the spectral based modellinghasdefinedafixedboundary.

10. DATA USE

the digital raster data is supplied without any post modellingorclassificationspatialfiltering.Datausersmaywish to consider applying spatial cleaning techniques, for example, such as applying a moving 3 x 3 pixel window filter toremove isolatedsingleclasspixels fromthedataset, especially if the raster data is to be converted to vector format.

11. METADATA

2013-14 GEOTERRAIMAGE SOUTH AFRICAN NATIONAL LAND-COVER DATASET DEA / CARDNO 35 x CLASS LEGEND:CORE METADATA ELEMENTS (SANS1878)

1(M) Dataset title: 2013-14 Gti Sa National land-Cover (Dea / CarDNo version) (dea_cardno_2014_sa_lcov_utm35n_vs2b_pivot-corr.img)

2(M) Dataset reference date: april 2013 - March 2014



3(O) Dataset responsible party: produced by Geoterra image (Gti) pty ltd, South africa

4(C) Geographic location of the dataset. MbR WestBoundlongitude: -717294.00 (Upper left x) eastBoundlongitude: 1301256.00 (lower right x) NorthBoundlongitude: -2239230.00 (Upper left y) SouthBoundlongitude: -4046670.00 (lower right y)

projection coordinates based on Universal transverse Mercator UtM 35 North, WGS84 (datum), meters.

5(M) Dataset language: “english” (eng)

6(C) Dataset character set: UtF8 (8-bit data)

7(M) Dataset topic category: 010 = Base Map earth coverage

8(O) Scale of the dataset: land-cover mapped from 30 metre resolution landsat satellite imagery, therefore recommended for ±1:75,000 - 1: 90,000 scale or coarse mapping & modelling applications.

9(M) Abstract describing the dataset: the 2013-14 South african National land-cover dataset produced by GeoterraiMaGe as a commercial data product has been generated from digital, multi-seasonal landsat 8 multispectral imagery, acquired between april 2013 and March 2014. in excess of 600 landsat images were used to generate the land-cover information, based on an average of 8 different seasonal image acquisition dates, within each of the 76 x image frames required to cover South africa. the land-cover dataset, which covers the whole of South africa, is presented in a map-corrected, raster format, based on 30x30m cells equivalent to the image resolution of the source landsat 8multi-spectral imagery.Thisspecificversionofthe2013-14 GeoterraiMaGe South african National land-Cover Dataset is supplied to the South african Department of environmental affairs under the Dea / CarDNo project: SCpF002:”implementation of land

Use Maps for South africa”. the supplied dataset referred to in this report represents asubset of the full 2013-14 GeoterraiMaGe South african National land-Cover Dataset, in terms of supplied land- cover / land-use categories. Use of this Dea / CarDNo data version is governed by the liCeNCe aGreeMeNt contained on this report. the Dea / CarDNo dataset contains 35 x land-cover / use information classes, covering a wide range of natural and man-made landscape characteristics. the original land-cover dataset was processed in UtM (north) / WGS84 map projection format based on the landsat 8 standard map projection format as provided by the USGS. the data remains the property of GeoterraiMaGe, and is protected by copyright laws. all intellectual property rights pertaining to the data remain with GeoterraiMaGe at all times.

This updated 2013-14 dataset corrects the coding error for DEA-CARDNOclasses#6and#7(cultivatedfieldsandpivotcultivatedfields)whichwasinadvertentlyswappedinthepreviousversion.Apartfromtherecodingofthetwofield-typeclasses, all other details and class codes etc remain exactly the same as the previously delivered datasets.

10(O) Dataset format name: erDaS imagine *img raster formats

11(O) Dataset format version: version01(file#22)

12(O) Additional extent information for the dataset: (vertical and temporal) vertical extent:

Minimum value: n/a

Maximum value: n/a

Unit of Measure: n/a

vertical Datum: n/a

temporal extent: land-cover datasets generated in January 2015, based on april 2013 - March 2014 multi-seasonal landsat 8 and 5 satellite imagery.

14. appendix G


14(O) Reference system: Universal transverse Mercator (UtM) 35 North

CrS:

projection Used: Universal transverse Mercator (UtM) 35 North

Spheroid used: WGS84

Datum used: WGS 84

ellipsoid parameters: ellipsoid semimajor axis axis units denominatorofflatteningratio

projection parameters: UtM Zone: 35 (North) Standard parallel

longitude of central meridian: 27:00:00.00 east latitude of projection origin: 00:00:00.00 east False easting: 500000.00 meters

False northing: 0.00 meters

Scale factor at equator: 0.999600

projection units: meters

15(O) Lineage statement: land-cover dataset generated in-house by Geoterraimage (pretoria) in January 2015, based on primarily multi-date landsat 8 imagery acquired between april 2013 and March 2014.

16(O) On-line resource: n/a

17(O) Metadata file identifier: n/a

18(O) Metadata standard name: SaNS i878

19(O) Metadata standard version: version 01

20(C) Metadata language: english (eng)

21(C) Metadata character set: 021 (Usascii)

22(M) Metadata point of contact: Name: Mark thompson

position Name: Director remote Sensing

organisation Name: Geoterraimage pty ltd

physical address:

Building Grain Building (1st Floor) Street WitheriteStreet_Suffix StreetStreet_Nr 477Suburb Die WilgersCity pretoriaZip 0041province Gauteng Country South africa

postal address:Box 295Suburb persequor parkCity pretoriaZip 0020province GautengCountry South africa

23(M) Metadata time stamp: 31 July 2015


APPENDIx A

2013-14 South African National Land-cover Legend (DEA / CARDNO legend)

the table below describes the 35 x land-cover / use classes contained in the Dea / CarDNo supplied subset of the 2013-14 South african National land-cover dataset, generated from multi-seasonal landsat 8 imagery.

See Appendix D for visual examples of percent cover densities per unit area.

PARENTDEA Class Name (and

digital code)Definition

Water Water (33)

areas of open, surface water, that are detectable on all image dates used in the landsat 8 based water modelling processes. the mapped extent represents the maximum detectable water extent from all available imagery acquired within the 2013-14 assessment period. includes both natural and man-made water features.

WetlaND Wetland (12)

Wetland areas that are primarily vegetated on a seasonal or permanentbasis.Definedonthebasisofseasonalimageidentifiablesurface vegetation patterns (not subsurface soil characteristics). Thevegetationcanbeeitherrootedorfloating.Wetlandsmaybeeither daily (i.e. coastal), temporarily, seasonal or permanently wet and/or saturated. vegetation is predominately herbaceous. includes but not limited to wetlands associated with seeps/springs, marshes, floodplains,lakes/pans,swamps,estuaries,andsomeriparianareas.Wetlandsassociatedwithriparianzonesrepresentimageidentifiedvegetation along the edges of watercourses that show similar spectral characteristics to nearby wetland vegetation. excludes Mangrove swamps. permanent or seasonal open water areas within thewetlandsareclassifiedseparately.Seasonalwetlandoccurrenceswithin commercially cultivated field boundaries are not shown,although they have been retained within subsistence level cultivation fields.

ForeSt indigenous Forest (1)

Natural / semi-natural indigenous forest, dominated by tall trees, where tree canopy heights are typically > ± 5m and tree canopy densities are typically > ± 75 %, often with multiple understory vegetation canopies. Note this class refers to Grassland areas not within the Fynbos, Nama Karoo and Succulent Karoo Biome boundaries (as defined in the SANBI 2006“Vegetation of Southafrica, Swaziland and lesotho” map data).

14. appendix G




tHiCKet & DeNSe BUSH

Dense Bush, thicket & tall Dense Shrubs (2)

Natural / semi-natural tree and / or bush dominated areas, where typically canopy heights are between 2 - 5m, and canopy density is typically > ± 75%, but may include localised sparser areas down to ± 60%22. includes dense bush, thicket, closed woodland, tall, dense shrubs, scrub forest and mangrove swamps. Can include self-seeded bushencroachmentareasifsufficientcanopydensity.Notethisclassrefers to Grassland areas not within the Fynbos, Nama Karoo and SucculentKarooBiomeboundaries(asdefinedintheSANBI2006“vegetation of South africa, Swaziland and lesotho” map data).

WooDlaND / opeN-BUSH

Woodland and open Bushland (3)

Natural / semi-natural tree and / or bush dominated areas, where typically canopy heights are between ± 2 - 5 m, and canopy densities typically between 40 - 75%, but may include localised sparser areas down to ± 15 - 20 %28. includes sparse - open bushland and woodland, including transitional wooded grassland areas. Can include self-seeded bush encroachment areas if canopy density is within indicated range. in the arid western regions (i.e. Northern Cape), this cover class may be associated with a transitional bush / shrub cover that is lower than typical open Bush / Woodland cover but higher and/or more dense than typical low Shrub cover. Note this class refers to Grassland areas not within the Fynbos, Nama Karoo andSucculentKarooBiomeboundaries (asdefined in theSANBI2006 “vegetation of South africa, Swaziland and lesotho” map data).

GraSSlaND Grassland (13)

Natural / semi-natural grass dominated areas, where typically the tree and / or bush canopy densities are typically < ± 20 %, but may include localised denser areas up to ± 40 %, (regardless of canopy heights)7. includes open grassland, and sparse bushland and woodland areas, including transitional wooded grasslands. May include planted pasture (i.e. grazing) if not irrigated. irrigated pastures will typically be classified as cultivated, and urban parks and golf courses etcunder urban. Note this class refers to Grassland areas not within the Fynbos, Nama Karoo and Succulent Karoo Biome boundaries (asdefinedintheSANBI2006“VegetationofSouthAfrica,Swazilandand lesotho” map data).

28Normallyitispreferredthatland-coverclassdefinitionsaremutuallyexclusiveintermsofcontentandassociatedlandscapecharacteristics.Due to the nature of the multi-seasonal spectral modelling approach used in compiling the land-cover dataset, it seems that there is some transitional overlap between “bushy / woody grassland” and “grassy woodland / bushland” in terms of woody cover densities. For this reasontheclassdefinitionsforboth“Woodland/OpenBush”and“Grassland”includebothaclassspecificcorewoodycoverdefinition,andasharedtransitionalwoodycoverdefinition.Thesamplepointsusedtoverifythemapaccuracyrecordtheobservedwoodycoverpercentage for these vegetation types should the user wish to undertake further analysis.




loW SHrUBlaND

low Shrubland (4)

Natural/semi-naturallowshrubdominatedareas,typicallywith≤2m canopy height. includes a range of canopy densities encompassing sparse to dense canopy covers. very sparse covers may be associated with the bare ground class. typically associated with low, woody shrub, karoo-type vegetation communities, although can also represent locally degraded vegetation areas where there is a significantlyreducedvegetationcoverincomparisontosurrounding,lessimpactedvegetationcover,includinglong-termwildfirescarsinsome mountainous areas in the western Cape. Note that taller tree / bush / shrub communities within this vegetation type are typically classifiedseparately as one of the other tree or bush dominated cover classes. Note this class refers to low Shrubland areas not within the Fynbos, Nama Karoo and Succulent Karoo Biome boundaries (as defined in theSANBI2006“VegetationofSouthAfrica,Swazilandand lesotho” map data).

CUltivateD

Commercial annuals (rainfed) (6)

Cultivated lands used primarily for the production of rain-fed, annual crops for commercial markets. typically represented by large fieldunits,often indense localorregionalclusters. Inmostcasesthe defined cultivated extent represents the actual cultivated orpotentially extent. includes Sugarcane crops.

Commercial pivot (7)

Cultivated lands used primarily for the production of centre pivot irrigated, annual crops for commercial markets. in most cases the defined cultivated extent represents the actual cultivated orpotentially extent. includes Sugarcane crops.

Commercial permanent (orchards) (8)

Cultivated lands used primarily for the production of both rain-fed and irrigated permanent orchard crops for commercial markets. includes both tree, shrub and non-woody crops, such as citrus, tea, coffee,grapes,lavenderandpineapplesetc.Inmostcasesthedefinedcultivated extent represents the actual cultivated or potentially extent.

Commercial permanent (vines) (9)

Cultivated lands used primarily for the production of both rain-fed and irrigated permanent vine (grape) crops for commercial markets. Inmost cases thedefinedcultivatedextent represents the actualcultivated or potentially extent.

Subsistence (10)

Cultivated lands used primarily for the production of rain-fed, annual crops for local markets and / or home use. typically represented by smallfieldunits,oftenindenselocalorregionalclusters.Thedefinedareamayincludeintra-fieldareasofnon-cultivatedland,whichmaybe degraded or use-impacted, if the individual field units are toosmalltobedefinedasseparatefeatures.

14. appendix G




ForeSt plaNtatioN

Forest plantations (5)

planted forestry plantations used for growing commercial timber tree species. the single class represents a combination of mature, young and temporary clearfelled stands. the class includes spatially smaller woodlots and windbreaks with the same cover characteristics.Notethatyoungsaplingsareverydifficulttoidentifyon 30 metre resolution landsat imagery if the actual tree canopy cover density is below ± 30 - 40%, because the background cover, for example, grassland, then dominates the spectral characteristics in that pixel area.

MiNe Mine (32)

Mining activity footprint, which includes extraction pits, tailings, wastedumps,floodedpitsandassociatedsurfaceinfrastructuresuchas roads and buildings (unless otherwise indicated), for both active and abandoned mining activities. Class may include open cast pits, sand mines, quarries and borrow pits etc.

Bare

Bare (Non vegetated) (34)

Bare, non-vegetated ground, with little or very sparse vegetation cover (i.e. typically < ± 5 - 10% vegetation cover), occurring as a result of either natural or man-induced processes. includes but not limited to natural rock exposures, dry river beds, dry pans, coastal dunes and beaches, sand and rocky desert areas, very sparse low shrublands and grasslands, erosion areas, and major road networks etc.Mayalsoincludelong-termwildfirescarsinsomemountainousareas in the western Cape.

Degraded (35)Sparsely vegetated areas, occurring as a result of man-induced processes, which show significantly lower overall vegetation covercomparedtosurrounding,naturalundisturbedareas.

BUilt-Up Settlements (11)

all built-up areas, represented as a single class, including but not limited to commercial, industrial, heath, education, religion, transport and residential land-uses, including both formal and informal structures, across a range of structural densities from high to low. i. high and low density areas containing high density buildings and other built-up structures associated with mainly non-residential, commercial, administrative, sport, health, religious or transport (i.e. train station) activities. includes agricultural smallholdings on the urban periphery.




FyNBoS

Fynbos: forest (14)Forest areas as per class (1) within the SaNBi Fynbos Biome boundary.

Fynbos: thicket (15)Dense Bush and thicket areas as per class (2) within the SaNBi Fynbos Biome boundary.

Fynbos: open bush (16)Woodland and open Bush areas as per class (3) within the SaNBi Fynbos Biome boundary.

Fynbos: low shrub (17)low Shrubland areas as per class (4) within the SaNBi Fynbos Biome boundary.

Fynbos: grassland (18)Grassland areas as per class (13) within the SaNBi Fynbos Biome boundary.

Fynbos: bare ground (19)Bare Ground areas as per class (14) within the SaNBi Fynbos Biome boundary.

NaMa Karoo

Nama Karoo: forest (20)Forest areas as per class (1) within the SaNBi Nama Karoo Biome boundary.

Nama Karoo: thicket (21)Dense Bush and thicket areas as per class (2) within the SaNBi Nama Karoo Biome boundary.

Nama Karoo: open bush (22)Woodland and open Bush areas as per class (3) within the SaNBi Nama Karoo Biome boundary.

Nama Karoo: low shrub (23)low Shrubland areas as per class (4) within the SaNBi Nama Karoo Biome boundary.

Nama Karoo: grassland (24)Grassland areas as per class (13) within the SaNBi Nama Karoo Biome boundary.

Nama Karoo: bare ground (25)

Bare Ground areas as per class (14) within the SaNBi Nama Karoo Biome boundary.

SUCCUleNt Karoo

Succulent Karoo: forest (26)Forest areas as per class (1) within the SaNBi Succulent Karoo Bi-ome boundary.

Succulent Karoo: thicket (27)Dense Bush and thicket areas as per class (2) within the SaNBi Succulent Karoo Biome boundary.

Succulent Karoo: open bush (28)

Woodland and open Bush areas as per class (3) within the SaNBi Succulent Karoo Biome boundary.

Succulent Karoo: low shrub (29)

low Shrubland areas as per class (4) within the SaNBi Succulent Karoo Biome boundary.

Succulent Karoo: grassland (30)

Grassland areas as per class (13) within the SaNBi Succulent Karoo Biome boundary.

Succulent Karoo: bare ground (31)

Bare Ground areas as per class (14) within the SaNBi Succulent Karoo Biome boundary.

14. appendix G


Note:theFynbos,NamaKarooandSucculentKaroovegetationsub-classesareprovidedwithadditionalclasscodesrepresentingtheparentprimaryvegetationclass,inorderthatdifferentlevelsofclassdetailcanbeusedandaccessedbydatausers.Thisisillustratedbelow. Dual legend system for the DEA CARDNO 2013-14 SA land-cover classes (statistics for pivot code corrected dataset).

Class Names (Color) represents the full 35 x land-cover / use class legend (with Biome level information)Class Names2 (Color2) represents the amalgamated 17 x land-cover / use class legend (no Biome level information)area refers to Hectares (Ha) and Histogram to the number of 30x30m image pixels.


APPENDIx b

List of accompanying documents and files.

excel spreadsheet containing full list of landsat 8 and 5 acquisition dates, per image frame, as used in the generation of the 2013-14 South african National land-Cover dataset29.

eSri arcGiS point coverage (UtM35 north) representing the sample points used to verify the 2013-14 South african National land-Cover dataset (version 4), and calculate the DEA-CARDNOlegendformat mapping accuracies.

29 Note that the listed acquisition dates represent the primary dates used in the land-cover modelling, and not any additional image dates that were only used for localised cloud masking on the primary image date. additional image dates are however listed if they themselves were also considered primary acquisition dates and used as modelling inputs.

14. appendix G


APPENDIx C

51 x Class Seasonally Defined Spectral Foundation Class Legend.

excel spreadsheet containing full list of landsat 8 and 5 acquisition dates, per image frame, as used in the generation of the 2013-14 South african National land-Cover dataset29.

eSri arcGiS point coverage (UtM35 north) representing the sample points used to verify the 2013-14 South african National land-Cover dataset (version 4), and calculate the DEA-CARDNOlegendformat mapping accuracies.


APPENDIx D

Visual representation of canopy cover density percentages per unit area.

adapted from paine, D.p., 1981, aerial photography and image interpretation for resource management: New york, John Wiley & Sons, 422 p.

14. appendix G


APPENDIx E

Contingency Matrix representing the Map Accuracy Results for all 33 x Land-Cover / Use Classes.


14. appendix G


14.2 1990 – 2013-14 Land use change report

1990 - 2013/14 South African National Land-Cover Change

Dea/CarDNo SCpF002: implementation of land-Use Maps for South africa

Project Specific Data Report

Data proDUCt CreateD ByGEOTERRAIMAGE (South Africa),

www.geoterraimage.comaugust 2015, version 01


LIAbILITY

the Client has to verify prior to his commitment, the adequacy of GeoterraiMaGe datasets and products to the specificdataandanalysesneedsassetoutbytheclient.GEOTERRAIMAGEwillinnoeventbeliableforanydirect,indirect,incidentalorconsequentialdamagesincludinglostsavingsorprofit,legalfees,lostdata,lostsales,lossofcommitments, business interruption, including loss of or other damage, resulting from its datasets and products, whether or not GeoterraiMaGe has been advised of the possibility of such damage. this includes damages incurred by the client or any third party.

GeoterraiMaGe datasets and products are processed from sources like topographic maps and satellite images. the quality of these sources is not the responsibility of GeoterraiMaGe. the data and data products are provided “as is”, without warranty of any kind, either express or implied, including, but not limited to, the implied warranties of merchantabilityandfitnessforaparticularpurpose.

Whilst all possible care and attention has been taken in the production of the data, neither GeoterraiMaGe or any of its respective sub-organisations or employees, accept any liability whatsoever for any perceived or actual inaccuracies or misrepresentations in the supplied data or accompanying report(s), as a result of the nature of mapping from coarse and medium scale satellite imagery. any mapped boundaries either implied or inferred within the data do not have any legal status whatsoever.

Allrightsnotspecificallygrantedorspecifiedinthisagreement,pertainingtothisdataarereservedtoGEOTERRAIMAGEpty ltd.

PLEASE NOTE

Thisreportprovidesinformationontheland-coverchangestatisticsderivedfromacomparisonofthespecificversionsof the 1990 and 2013/14 GeoterraiMaGe South african National land-Cover Dataset supplied to the South african Department of environmental affairs under the Dea / CarDNo project: SCpF002:”implementation of land Use Maps for South africa”. these datasets represent information subsets of the full 1990 and 2013/14 GeoterraiMaGe South african National land-Cover Datasets, in terms of supplied land-cover / land-use categories and associated class legend.

14. appendix G


CONTENTS

1. BaCKGroUND 388

2. CoMMeNtS oN 1990 - 2014 Sa laND-Cover CHaNGe aSSeSSMeNt 388

2.1InfluenceofLand-CoverMappingAccuracies 388

2.2InfluenceofSeasonal/ClimaticDifferencesonLand-CoverChange 388

2.3InfluenceofMappingDetailonLand-CoverChange 391

3. QUalitative exaMpleS oF laND-Cover CHaNGe 1990 - 2013/14 391

4. QUaNtitative aSSeSSMeNt oF laND-Cover CHaNGe 1990 - 2013/14 397

4.1 National level land-Cover Change results 397

4.2 provincial level land-Cover Change results 399

4.3 inter-Class land-Cover Change results 402

5. 1990 - 2013/14 CHaNGe reSUltS aNalySiS 409

5.1 indigenous Forest 409

5.2 thicket / Dense Bush 409

5.3 Woodland/ open Bush 409

5.4 low Shrubland 409

5.5 plantations / Woodlots 410

5.6 Cultivated Commercial annual Crops Non-pivot 410

5.7 Cultivated Commercial annual Crops pivots 410

5.8 Cultivated Commercial permanent orchards 410

5.9 Cultivated Commercial permanent vines 410

5.10 Cultivated Subsistence Crops 411

5.11 Settlements (including Smallholdings) 411

5.12 Wetlands 412

5.13 Grasslands 412

5.14 Mines 413

5.15 Waterbodies 413

5.16 Bare Ground 413

5.17 Degraded 413


1. bACKROUND

an assessment of land-cover change has been undertaken using the 1990 and 2013-14 South african National land-Cover Datasets. Both of these dataset have been generated from comparable historical and current landsat satellite imagery, using the same model-based mapping techniques.

the two source datasets are therefore, as far as possible, directly comparable in terms of land-cover content, spatial detail and associated mapping accuracies. the production of both of these national land-cover datasets is fully documented in separate reports pertaining to each land-cover dataset. It is assumed that a user of the statistical land-cover change data is fully aware and appreciative of the information contained in these land-cover data generation reports.

in preparation for the land-cover change assessment, the original 35 x class national land-cover datasets were re-formattedintotheDEArequested,simplified17xclasslegend format to be used for land-cover change reporting.

the tables (1 & 2) illustrate the class and code conversions that were used to collapse the 35 x class legend into the 17 x class legend.

2. COMMENTS ON 1990 - 2014 SA LAND-COVER CHANGE ASSESSMENT

Therearethreebasicfactorsthatmayhaveinfluencedland-cover change results and associated statistics, namely land-cover mapping accuracies, seasonal / climatic conditions under which the original source imagery was acquired, and the level of (landscape) detail that is being compared over time.

2.1 Influence of Mapping Accuracies on Land-Cover Change

Mapping accuracies have been calculated for the 2013-14 land-cover dataset and are reported in the 2013-14 Data User’s report. the results of the 2013-14 map accuracy assessment are assumed to be representative of the 1990 accuracies, since comparable image data and modelling / mapping techniques were used for both datasets. No independent map accuracy assessment was undertaken for the 1990 dataset due to the unavailability of suitable reference data. the reliability of (land-cover) changestatisticsaregreatlyinfluencedbytheaccuracyof the input data against which change is determined. in simplistic terms, the sum of the known mapping error associated with a particular land-cover class over two time periods can indicate the level of ‘statistical change’ that is in fact a result of compounded mapping errors over time, and not actual land-cover change. For this reason all statistical (land-cover) change values must be interpreted and evaluated in the context of the reported land-cover class mapping accuracy, especially the User (as opposed to theProducers)accuracy.Classspecificmappingaccuraciesare included in the land-cover change results reported below and in the accompanying spreadsheet.

2.2 Influence of Seasonal / Climatic Differences on Land-Cover Change

Itisimportanttounderstandtheinfluencethatseasonal,and possible climatic differences between 1990 and 2013-14 may have had on the generation of each land-cover dataset in terms of landscape characteristics, and the possible effect of such differences on the change assessment results and associated spatial statistics.

the seasonal representation and range of monthly acquisition dates is generally better for the 1990 land-cover dataset, than used for the 2013-14 dataset, despite the same average number of acquisition dates per frame for the two datasets. this is because the 2013-14 dataset was limited to what cloud free imagery was recorded

14. appendix G


14. appendix G


between april 2013 and the designated March 2014 cut-off, whereas the 1990 dataset had access to a wider range of imagery from 1989 - 1993. this greater seasonal representationmayhaveinfluencedthemodellingresultsin terms of the distribution and extent of some of the natural vegetation classes (excluding indigenous forests), sinceabetterseasonalprofilewasoftenpossiblein1990compared to 2014 in some image frame locations.

the 1990 period appears to have been generally wetter than the 2013-14 period in most regions, as observed through the increase in observable surface water features in 1990. this is especially evident in the pans across the central regions of the Free State and Northern Cape, wheremanypansarewaterfilledin1990anddryin2013-14.Thesedifferenceswillreflectaschangesbetweenthetwo assessment periods but do not necessarily represent a permanent loss of water bodies over the ± 24 year period, but rather a seasonal (or climatic) induced difference. this fact needs to be noted during modelling and the interpretation of land-cover change results and associated statistics.

the wetter conditions prevalent in the 1990 dataset have allowed the wetland modelling to be extended across the full national coverage, unlike the drier 2013-14 image data which restricted wetland modelling to the wetter eastern andsouthernregions.Thesedifferenceswillreflectaschanges between the two assessment periods but do not necessarily represent a permanent loss of wetlands over the ± 24 year period, but rather a seasonal (or climatic) induced difference. this fact needs to be noted during modelling and interpretation of land-cover change results.

the wetter conditions prevalent in the 1990 dataset mayhaveinfluencedtheboundarybetweenlow-shrub(i.e. karoo type) and highveld grassland vegetation, since the 1990 grassland / low shrub transition in the western Free State region appears to have shifted generally to the west compared to 2013-14, which may be a result of a greater grass component being present in the karoo / grass transition zone in 1990.

the (highveld, montane and coastal) grassland correction procedures applied to the 1990 dataset are considered to be slightly more accurate than those applied to the 2013-14data,asaresultofbothmodifiedproceduresusedand better image seasonal representation in 1990. it is therefore potentially possible that some of the ‘grassland’ to ‘open woody vegetation’ changes in these regions may be a result of the improved grassland delineation in 1990 and not actual bush cover increases in 2013-14.

2.3 Influence of Mapping Detail on Land- Cover Change

the ability to determine change is obviously dependent on the level of detail that is being observed and compared. For example, changes in the extent of urban areas between 1990and2013-14maynotbeas locallysignificantasmaybeexpected,sincethetotalbuilt-upfootprintdefinedin the “settlements” class in both the 1990 and 2013-14 datasets includes peripheral smallholding areas around the mainbuilt-upareas;andthesetendtobethefirstland-usethat is converted to formal urban areas, before further expansion into natural and cultivated lands.

3. QUALITATIVE ExAMPLES OF LAND-COVER CHANGE 1990 - 2013/14.

the following screen-shots provide visual examples of the type of land-cover changes that have been observed and mapped over the ± 24 year period between 1990 and 2013/14, based on the landsat derived land-cover datasets. 1990 land-cover is on the leFt and 2013/14 land-cover is on the riGHt in all examples.


Orange / Vaal River Confluence, Douglas, Northern Cape

Notesignificantincreaseincentrepivotirrigationfieldsin2013/14comparedto1990.

Villages, Giyani, Limpopo

Note the bush clearing effects around villages in 2013/14 compared to 1990, and new dam in 2013/14.

14. appendix G


Kleinmond, Western Cape

Note new settlements and reduction of plantations (orange) in 2013/14 compared to 1990. the light green and pink areasin2013/14arelowvegetationareasarelikelytobetheresultofmountainfirescarsthatwerenotevidentin1990.

Kosi bay, KwaZulu Natal

Note the increase in plantation forestry (orange) in 2013/14 and the greater extent of wetlands (light blue) in 1990 as a result of the (generally) wetter conditions.


KZN Midlands, KwaZulu Natal

Note the increase in plantation forestry (orange) in 2013/14.

Midrand, Gauteng

Notesignificantexpansionofsettlements(yellow)in2013/14comparedto1990.Notethatthesettlementclassincludessmallholdings, so in this area the expansion of settlements has gone into other land-cover classes as well, such as natural grasslands.

14. appendix G


Pan (wetlands), Northern Cape

Notesignificantlygreaterextentandnumberofwaterfeaturesevidentinthepansin1990asresultofthewetterconditions compared to 2013/14.

Olifants River, Western Cape

Notesignificantincreaseincentrepivotirrigationfieldsin2013/14comparedto1990.


Rustenburg, North West

Notethesignificantincreaseinminingactivity(red)in2013/14comparedto1990.

Witbank / eMalahleni, Mpumalanga

Notethesignificantincreaseinminingactivity(red)in2013/14comparedto1990,andtheurbanexpansion.

14. appendix G


4. QUANTITATIVE ASSESSMENT OF LAND-COVER CHANGE 1990 - 2013/14.

a quantitative assessment of land-cover change between 1990 and 2013/14 has been completed, with the results presented below in terms of both change reported in terms of the original 30m resolution raster cell format, and also in terms of a spatially re-sampled 90x90 cell format.

the re-sampled 90m cell format data was created by a spatial aggregation process that determined the majority land-cover from a window of 3 x 3 original 30m cells and returned this to a new, single 90m raster cell representing the equivalent spatial area. Comparison of (national) change statistics derived from both the original

30m resolution raster cell data and the re-sampled 90m resolution cell data would provide an indication of whether ornotasignificantlevelofvariationwasintroducedintothe change results from pixel level mapping anomalies compared to the local majority land-cover characteristics.

4.1 National Level Land-Cover Change Results

the tables below show the results of comparing the 1990 land-cover data to the 2013/14 land-cover, on a national basis, based on both the original 30m raster format data and the spatially re-sampled 90m raster cell format data, which represents the majority land-cover class in a ± 1 ha area. Users are reminded again to interpret all statistical (land-cover) change values in the context of the reported land-cover class mapping accuracy (especially

St. Lucia, KwaZulu Natal

Notethereductionofplantations(orange)ontheeasternshoresofStLuciain2013/14andthesignificantsettlementexpansion (yellow) and loss of indigenous forest (dark green) in Dukuduku Forest in 2013/14, in the southern central section. the wetter conditions in 1990 have resulted in more observable wetlands (light blue) compared to 2013/14.


Users accuracy), and overall number of mapped image pixels.Classspecificmappingaccuraciesareincludedinthe land-cover change results reported below and in the accompanying spreadsheet.

the percentage change results for all classes show very similar values in both the original 30m and re-sampled

(3) National Change Statistics 1990 - 2013/14, based on original 30m raster data.

90m datasets, which indicates that the comparison of the original30mdatasetsisnotintroducinganysignificantdifference as a result of single pixel anomalies. the 2013/14 land-cover mapping accuracies have been included in the tables to allow the change results to be assessed in terms of the class mapping accuracies achieved, so that classes.

(4) National Change Statistics 1990 - 2013/14, based on re-sampled 90m raster data.

14. appendix G


4.2 Provincial Level Land-Cover Change Results

the change statistics for each province are shown below. these have been generated from the original 30m resolution format data, and show regional trends in land-

cover / land-use change over the ±24 year period between 1990 and 2013/14. the 2013/14 land-cover mapping accuracies have been included in the tables (5 - 13) below to allow the provincial change results to be assessed in terms of the associated class mapping accuracies.


14. appendix G


4.3 Inter-Class Land-Cover Change Results

the tables below illustrate the inter-class changes between 1990 and 2013/14, and allow an interpretation of (a) what the 1990 extent of land-cover class x has changed to in 2013/14, or (b) what the current extent of class y in 2013/14 was, in terms of 1990 land-cover characteristics. these statistics are based on the fact that a 17 x class land-cover legend in both 1990 and 2013/14 means that there are 289 possible solutions to what a 1990 class can become in 2013/14. interpretation of these inter-class changes should again be assessed in terms of the associated class mapping accuracies. Note that the table statistics are based on the original 30m resolution datasets.

the table (14) on the right shows the numeric codes used to represent all possible 289 inter-class changes between 1990 and 2013/14.

Thetable(15)ontheoppositepageprovidesthedefinitionof each of the 289 inter-class changes, in terms of what 1990 class changed to what 2013/14 class, together with the number of 30m resolution pixel (not ha) occurrences.

14. appendix G


14. appendix G


the table (16) describes the actual 1990 - 2013/14 inter-class land-cover changes for the whole of South africa, in terms of Hectares (Ha):

14. appendix G


the table (17) describes the actual 1990 - 2013/14 inter-class land-cover changes for the whole of South africa, expressed in terms of the percentage of the original 1990 class extent and what it has become in 2013/14. For example, 81.84 % of what was mapped as indigenous forest in 1990 is still indigenous forest in 2013/14, and 11.85% of what was indigenous forest in 1990 is now mapped as thicket / dense bush in 2013/14:


the table (18) describes the actual 1990 - 2013/14 inter-class land-cover changes for the whole of South africa, expressed in terms of the percentage of the current 2013/14 class extent and what it was originally in 1990. For example, 71.95 % of what was mapped as indigenous forest in 2013/14 was also indigenous forest in 1990, and 11.26% of what is indigenous forest in 2013/14 was mapped as thicket / dense bush in 1990:

14. appendix G


5. 1990 - 2013/14 CHANGE RESULTS ANALYSIS

5.1 Indigenous Forest

according to the land-cover data there was nationally a 13.75% increase in the indigenous forest class extent between 1990 and 2013/14. However the 2013/14 map accuracy results indicate a 72.6 % Users accuracy, which may mean that in many cases the reported change is a result of mapping error between true forest and the spectrally similar woody vegetation classes such as ‘thicket / dense bush’, as indicated in the 2013/14 accuracy report. in Gauteng province there is a 1350% (!) statistical increase in the extent of indigenous forest in 2013/14, but this is simply due to the fact that in 1990 2 pixels were mapped as forest whereas in 2013/14 29 pixels were mapped. this difference is likely to be a result of the reported mapping inaccuracies; with 29 pixels representing only 0.0006 % of the total mapped area of indigenous forest in 2013/14. National and provincial statistics should not however detract from the very relevant and real changes in indigenous forest cover change that are contained within the land-cover datasets in a local basis, such as around Dukuduku, St lucia (see previous picture example).

5.2 Thicket / Dense bush

according to the land-cover data there was nationally a 24.76% increase in the thicket / dense bush class extent between 1990 and 2013/14. However the 2013/14 map accuracy results indicate only a 53.74 % Users accuracy, which may mean that in many cases the reported change will be a result of mapping error between this class and other spectrally similar woody vegetation classes, as indicated in the 2013/14 accuracy report. provincial changes, in the woody dominated areas vary from +16.03% in limpopo, +18.77% in KZN, +36.46% in Mpumalanga, +79.52% in eastern Cape and -25.57 % in North West. it is recommended that such statistical changes are treated with caution unless validated by corroborating evidence

of plausible local change within the actual land-cover map datasets.

5.3 Woodland/ Open bush

according to the land-cover data there was nationally a 12.96% increase in the woodland / open bush class extent between 1990 and 2013/14. However the 2013/14 map accuracy results indicate only a 60.84 % Users accuracy, which may mean that in many cases the reported change will be a result of mapping error between this class and other spectrally similar woody vegetation classes, as indicated in the 2013/14 accuracy report. provincial changes, in the woody dominated areas vary from +3.47% in limpopo, +20.57% in KZN, - 4.73% in Mpumalanga, +115.79% in eastern Cape and +8.07% in North West. it is recommended that such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.

5.4 Low Shrubland

according to the land-cover data there was nationally a 1.67% increase in the low shrub class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 70.59 % Users accuracy, which may mean that the limited amount of change is a function of both increasing map accuracy and the extremely large national class footprint withinwhichanylocalchangeisinsignificantintermsoftotal area affected. provincial changes low shrub / karoo / Fynbos dominated areas vary from +48.59% in Free State, -2.45% in Northern Cape and +7.23% in Western Cape. in most cases the changes in low shrub extent seem to be along a transitional gradient with grasslands, and may bereflectinglocalannualchangesingrasscomponentcover in response to different annual rainfall conditions, as is evident in the difference in open water / wetland areas between 1990 and 2013/14. it is recommended that such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.


5.5 Plantations / Woodlots

according to the land-cover data there was nationally a 2.55% decrease in the plantation / woodlot class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 89.3 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. the nationalstatisticdoesnothoweverreflectprovincialor local area changes over the same ± 24 year period, withsomeareasdefinitelyshowinglossesandothersexpansion. provincial changes in traditional forestry areas vary from -23.93% in limpopo, +1.1% in Mpumalanga, +1.02% in eastern Cape and +3.57% in KZN.

5.6 Cultivated Commercial Annual Crops Non-Pivot

according to the land-cover data there was nationally a 7.62% decrease in the cultivated commercial annual crop non-pivot class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 91.91 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. the national statistic does not however reflectprovincialorlocalareachangesoverthesame±24yearperiod,withsomeareasdefinitelyshowinglossesand others expansion. provincial changes in traditional forestry areas vary from -20.25% in limpopo, -12.65% in Mpumalanga, +11.64% in KZN, -10.30% in eastern Cape and -3.33% in Western Cape for example. on a local scale, many of these losses and expansions also appear to relate to associated increases in other cultivation land-uses, such as orchards, vines and pivots, and losses due to new mining and urban areas.

5.7 Cultivated Commercial Annual Crops Pivots

according to the land-cover data there was nationally a 220.16% increase in the cultivated commercial annual crop pivot class extent between 1990 and 2013/14. this is

by far the biggest increase in any land-cover class extent over the ±24 year period and represents a huge increase in irrigation activities and associated water consumption. the 2013/14 map accuracy results indicate a 95.381 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and very likely to be accurate. the national statistic does not howeverreflectprovincialorlocalareachangesoverthesame ± 24 year period, with provincial changes in pivot class extent varying from +113.87% in limpopo, +262.41% in Mpumalanga, +261.41% in KZN, +413.66% in eastern Cape and +290.11% in Western Cape for example. all provinces showed increases in pivot class extent since 1990. on a local scale this expansion appears to have resulted in losses to both other cultivated land-uses and natural vegetation types, depending on location.

5.8 Cultivated Commercial Permanent Orchards

according to the land-cover data there was nationally a 10.64% increase in the cultivated commercial permanent orchard class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 92.18 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. the national statistic does not however reflectprovincialorlocalareachangesoverthesame±24yearperiod,withsomeareasdefinitelyshowinglossesand others expansion. provincial changes in traditional orchard areas vary from +40.27% in limpopo, +35.38% in Mpumalanga, +28.56% in Northern Cape, and -18.18 in eastern Cape for example. on a local scale, many of these losses and expansions also appear to relate to associated increases in other cultivation land-uses, such as orchards, vines and pivots, and losses due to new mining and urban areas.

5.9 Cultivated Commercial Permanent Vines

according to the land-cover data there was nationally a

14. appendix G


16.23% increase in the cultivated commercial permanent vines class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 91.61% Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate.Thenationalstatisticdoesreflectprovinciallevel changes over the same ± 24 year period, with similar percent increases being reported in both the Western and Northern Cape. all provinces with vines in 1990 show an increase in 2013/14.

5.10 Cultivated Subsistence Crops

according to the land-cover data there was nationally only a 2.83% increase in the cultivated subsistence crop class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 89.00 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. it is important to note however that unlike commercialfieldswhichwere(originally)mappedonanindividualfieldunitbasis,subsistencefieldsweremappedaccording to the outer extent of clusters of subsistence fields,asnecessitatedbytheircharacteristicallysmallareas compared to image resolution. this resulted in varieslevelsandpatternsofinter-fieldnon-croplandbeing included in the subsistence cluster boundaries. Duringtheland-covermappingtheseinter-fieldnon-crop areas have been re-classified, where possible, into their basic vegetation cover types, i.e. grass, dense bush etc. Since this was based on a generic modelling procedureasopposedtoimage-specificclassificationroutines, it is possible that the extent and coverage ofsuchinter-fieldnon-cropareashasbeeninfluencedby (a) the accuracies associated with these vegetation classes, and (b) any modifying effect resulting from 1990 conditions appearing to be generally wetter than those in 2013/14. Furthermore, the actual intensity of land-use (i.e. subsistence crop yield) cannot be gauged from the single ‘subsistenceclass’definedinthe17xclasslegendformat,sinceallsubsistencefieldshavebeenincludedwithinoneamalgamated class on the basis of (image) observable

fieldboundaries,regardlessofwhethersuchfieldswerecultivated, temporary or long term fallow (i.e. abandoned) in the assessment year. a subdivision of land-use activity intensity (a.k.a. cropping), based on annual Normalised DifferenceVegetationIndex(NDVI)profilesisavailableas part of the associated 72 x class Dea / Gti national land-cover dataset for 2013/14, but not presently for 1990. this may be a consideration for future work and analysis since it will allow the historical and current extents of high intensity subsistence cropping units to be compared, as opposed to just the overall extent of image observable fieldclusterboundaries.

Thenationalstatisticdoesnothoweverreflectprovincialor local area changes over the same ± 24 year period, with areas showing losses and others expansion. provincial changes in subsistence cropping vary from -13.34% in limpopo, -27.27% in Mpumalanga, and +30.29% in KZN. . it is recommended that until such time that that subsistence changes can be validated in terms of land-use intensity changes (i.e. productive-to-productive, productive-to-fallow,, fallow-to-productive etc), such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.

5.11 Settlements (including Smallholdings)

according to the land-cover data there was nationally only a 6.03% increase in the settlement class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 93.90 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. the national statisticdoesnothoweverreflectprovincialorlocalareachanges over the same ± 24 year period, with some areas definitelyshowinglossesandothersexpansion,rangingfrom +28.82% in limpopo to (surprisingly) -2.93% in KZN. Gauteng only shows a +10.58% increase, which from an anecdotal perspective may be lower than expected, but this is potentially explained by the fact that the settlement footprintdefinedinthetwoland-coverdatasetsincludes


a variety of urban sub-classes, including commercial, residential, village, schools etc, as well as smallholdings. peripheral smallholdings around urban centres are typicallythefirstland-useclasstoconverttoformalurbanareas under urban expansion. thus the total settlement footprintasdefinedintermsoftheDEA/CARDNOnational land-cover datasets will not change over time when smallholdings are converted to formal residential or industrialareas.Settlementsub-classesdefiningseparatecommercial, industrial, residential, informal, village and smallholding categories are available within the full 72 x class Dea / Gti national land-cover dataset for 2013/14, but not presently for 1990. this may be a consideration for future work and analysis since it will allow a more representative interpretation of urban expansion to be evaluated.

5.12 Wetlands

according to the land-cover data there was nationally a 32.78% decrease in the wetland class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 88.07 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. However wetlands is onespecificland-coverclassthathasbeensignificantlyaffected by the general wetter landscape conditions in 1990 compared to 2013/14, especially in the more conventionally drier western areas. the generally wetter landscape conditions in 1990 allowed wetland modelling to be undertaken across the entire country, whereas in 2013/14, because of drier conditions, wetland modelling was restricted to the southern and eastern regions. thus any wetlands in the normally drier western regions that were mapped in 1990 have not been re-mapped in 2013/14 simply because they were not visible as such on the imagery under these drier conditions. this does not mean that such wetlands have been ‘lost’ but rather are temporary absent in terms of surface representation in 2013/14. this must be noted when interpreting the statistics. a possible solution may be to take the combined area of wetlands from 1990 and 2013/14 and subtract

appropriate non-wetland land-use class extents from this for both 1990 and 2013/14 landscapes in order to derive a more accurate picture of wetland loss, irrespective of whether the wetlands

Thenationalstatisticdoesnothoweverreflectprovincialor local area changes over the same ± 24 year period, with losses ranging from -63.24% in Northern Cape to -12.13% in Gauteng, where the Northern Cape difference is likely to be a result of the landscape wetness differences already discussed, and the Gauteng difference probably being more representative of actual loss. all provinces showed a loss in wetland extent between 1990 and 2013/14. it is therefore recommended that such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.

5.13 Grasslands

according to the land-cover data there was nationally a 6.17% decrease in the grassland class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 84.56% Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. as indicated in the 1990 Dea / CarDNo Sa land-cover data report the post-modelling grassland correction edits are potentially more accurate than those applied to the 2013/14 land-cover dataset,asaresultofbetterimagedatesandmodifiedcorrection procedures. this may mean that the actual grassland extent as represented in the 1990 dataset is slightly more accurate than that represented in the 2013/14 dataset, and that there could be a corresponding increase in the level of grassland / woody vegetation cover confusion in 2013/14. it is recommended that such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.

the generally wetter landscape conditions in 1990, especially in the more typically arid karoo regions may

14. appendix G


also have experienced a regional annual increase in the grass component in the low shrub / grassland transition zone, also increasing the extent of the grassland class mapped in 1990.

Thenationalstatisticdoesnothoweverreflectprovincialor local area changes over the same ± 24 year period, with some areas showing losses and others expansion. provincial changes vary from -28.33% in limpopo, +1.93% in Mpumalanga, and +36.94% in Northern Cape.

5.14 Mines

according to the land-cover data there was nationally a 12.76% increase in the mining class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 92.82% Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and very likely to be accurate. the national statisticdoesnothoweverreflectprovincialorlocalareachanges over the same ± 24 year period, with both losses and increases being recorded, ranging from -46.22% in eastern Cape to +66.84% in Mpumalanga (as would be expected due to the expansion of open cast coal mining in this area). Since mapping accuracies are high for this class, interpretation of the statistical changes must be made in the context of the original and current spatial extents.

5.15 Waterbodies

according to the land-cover data there was nationally a -7.10% decrease in the water class extent between 1990 and 2013/14, which includes both natural, man-made, seasonal and permanent water features. the 2013/14 map accuracy results indicate a 79.64% Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. the national statistic does not however reflectprovincialorlocalareachangesoverthesame± 24 year period, with both losses and increases being recorded, ranging from +25.34% in limpopo, 11.35% in

Mpumalanga, and -59.75% in Northern Cape. these regionaldifferencesarelikelytohavebeeninfluencedby a combination of both the generally wetter landscape conditions in 1990, especially in the typically drier western regions, and by the construction of several new major dams after 1990 that are observable in 2013/14. Since mapping accuracies are high for this class, interpretation of the statistical changes must be made in the context of the original and current spatial extents, so that small spatial features do not become over-represented statistically if theyhaveshownsignificantstatisticalbutnotnecessarilyspatial growth.

5.16 bare Ground

according to the land-cover data there was nationally a -6.07% decrease in the bare ground class extent between 1990 and 2013/14. the 2013/14 map accuracy results indicate a 98.77% Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and very likely to be accurate. the national statisticdoesnothoweverreflectprovincialorlocalareachanges over the same ± 24 year period, with both losses and increases being recorded, ranging from +635.08% in limpopo, 2121.80% in North West and -63.01% in eastern Cape. Since mapping accuracies are high for this class, interpretation of the statistical changes must be made in the context of the original and current spatial extents in any given area, so that small spatial features do not become over-represented statistically if they have shown significantstatisticalbutnotnecessarilyspatialgrowth.The2013/14datasetmayhavebeeninfluencedbythegenerally drier conditions in some areas, perhaps resulting in lower vegetative cover, although in some regions human activity will have increased the bare ground extent.

5.17 Degraded

according to the land-cover data there was nationally a -36.61% decrease in the degraded class extent between 1990 and 2013/14. the 2013/14 map accuracy results


indicate a 78.69 % Users accuracy, which should mean that both the mapped and statistical representations of change are plausible and likely to be accurate. However, degradedisdefinedasareasshowingsignificantlowertotal vegetation cover to surrounding non-impacted areas, in terms of the viewed landscape. it is feasible therefore thatasignificantportionofthedifferenceindegradedareas between 1990 and 2013/14 could be attributed to the generally wetter conditions in 1990 compared to the drier 2013/14 conditions. the provincial results show asignificantlyhigherareaofbaregroundin1990than2013/14, apart from North West. this would appear to be the opposite of what would be expected give that 1990 was the wetter period, and thus the vegetation cover would be assumed to be more extensive. However an alternative suggestion maybe that, in terms of the generic modellingproceduresusedtodefinethisclass,whenallthe vegetation is generally lower (as in a drier period) then the difference between impacted (i.e. degraded) land and non-impacted land is harder to determine; compared to when non-impacted land is much greener in wetter conditions.

it is recommended that such statistical changes are treated with caution unless validated by corroborating evidence of plausible local change within the actual land-cover map datasets.

ADDITIONAL INFORMATION

PLEASE NOTE THAT THIS IS NOT A LAND-COVER MAP DATASET bUT A SPATIAL REPRESENTATION OF 1990 TO 2013/14 LAND-COVER CHANGE ON A PER-PIxEL bASIS.

14. appendix G

www.environment.gov.za

environment House473 Steve Bikocnr Steve Biko and Soutpansberg roadarcadiapretoria, 0083South africa

postal addressDepartment of environmental affairsp o Box 447pretoria0001

publishing date: august 2017

GHG NatioNal iNveNtory report South africa · 2018-11-07 · 6 GHG National inventory report for...

Documents

Transcript of GHG NatioNal iNveNtory report South africa · 2018-11-07 · 6 GHG National inventory report for...